Information processing device, information processing method, and program product

Abstract

The invention relates to an information processing apparatus, an information processing method, and a program product. Provided is an information processing apparatus. The method includes a step of setting a plurality of time windows for cutting data from raw data by setting two physical quantities related to damage of a device or a member as a first feature quantity and a second feature quantity, and a step of cutting data from the raw data. A step for calculating a frequency distribution in the raw data and a frequency distribution in the extracted data with respect to a plurality of first feature quantities divided by the second feature quantities; and a step for determining whether or not the raw data and the extracted data are similar using the frequency distribution. The information processing device repeatedly executes these steps by changing the settings of the plurality of time windows, and outputs extracted data similar to the original data.

Description

Technical Field

[0001] This disclosure relates to an information processing apparatus, information processing method, and program product. Background Technology

[0002] Japanese Patent Application Publication No. 2008-108247 discloses an information processing apparatus that reduces the size of the data used for analysis by compressing the raw data. The raw data used for analysis is data collected within a predetermined period using sensors mounted on a vehicle.

[0003] The information processing device disclosed in the aforementioned patent document compresses data by extracting data obtained at the time point when a certain vehicle speed is reached and data obtained at the time point when the vehicle speed inflection point from the original data.

[0004] The device can analyze damage to equipment or components mounted on the vehicle using data acquired from multiple sensors mounted on the vehicle. If data suitable for analyzing the extent of damage can be extracted from the raw data, the extent of damage can be analyzed in a shorter time compared to using the raw data. The aforementioned information processing device uses vehicle speed, location information, and driving data at the time of travel. This device correlates vehicle speed with driving position to extract vehicle speed change patterns and sets a driving schedule for the engine and motor to minimize fuel consumption. However, the aforementioned information processing device cannot extract data suitable for analyzing the extent of damage to equipment or components mounted on the vehicle from the raw data. Summary of the Invention

[0005] According to one aspect of this disclosure, an information processing apparatus is provided, configured to extract a portion of raw data collected within a predetermined period using multiple sensors mounted on a vehicle to analyze the degree of damage to equipment or components mounted on the vehicle. The information processing apparatus includes a processing circuit configured to perform a first process, wherein the first process designates a physical quantity in the raw data related to damage to the equipment or component as a first feature quantity, designates a physical quantity in the raw data that differs from the first feature quantity as a second feature quantity, divides the raw data into multiple datasets using the second feature quantity, and calculates the frequency distribution of the first feature quantity in each of the multiple datasets obtained from the division. The processing circuit is configured to modify the data during the period for extracting a portion of the raw data. The following processes are repeatedly performed using multiple time windows: a second process, setting the multiple time windows such that the sum of the periods of all time windows is shorter than the predetermined period; a third process, extracting data from the original data through the multiple time windows; a fourth process, corresponding to the division of the original data into multiple datasets, dividing the extracted data obtained by combining all the data extracted through the multiple time windows into multiple datasets, and calculating the frequency distribution of the first feature in each of the multiple datasets; and a fifth process, using the frequency distribution to determine whether the original data and the extracted data are similar, wherein the processing circuit is configured to use the extracted data similar to the original data to analyze the degree of damage to the device or the component based on the first feature and the second feature.

[0006] According to one aspect of this disclosure, an information processing method is provided, in which a processing circuit extracts a portion of data from raw data collected by multiple sensors mounted on a vehicle over a predetermined period to analyze the degree of damage to equipment or components mounted on the vehicle, wherein...

[0007] The information processing method includes the following steps: the processing circuit performs a first processing step, wherein the first processing step sets a physical quantity related to the damage of the device or the component contained in the original data as a first feature quantity, sets a physical quantity in the original data that is different from the first feature quantity as a second feature quantity, divides the original data into multiple datasets using the second feature quantity, and calculates the frequency distribution of the first feature quantity in each of the multiple datasets obtained by division; the processing circuit changes the settings of multiple time windows used to extract a portion of the original data to repeatedly perform the following processing step: a second processing step, wherein the period obtained by summing the periods of all time windows is shorter than the first feature quantity. The process involves setting multiple time windows within a predetermined period; a third process, extracting data from the original data through the multiple time windows; a fourth process, corresponding to the division of the original data into multiple datasets, dividing the extracted data obtained by combining all the data extracted through the multiple time windows into multiple datasets, and calculating the frequency distribution of the first feature in each of the multiple datasets; a fifth process, using the frequency distribution to determine whether the original data and the extracted data are similar; and the processing circuit using the extracted data similar to the original data to analyze the degree of damage to the device or the component based on the first feature and the second feature.

[0008] According to one aspect of this disclosure, a program product is provided that enables a processing circuit to extract a portion of raw data collected using multiple sensors mounted on a vehicle over a predetermined period to analyze the degree of damage to equipment or components mounted on the vehicle. The program product causes the processing circuit to perform a first process, which designates a physical quantity in the raw data related to damage to the equipment or component as a first feature quantity, designates a physical quantity in the raw data that differs from the first feature quantity as a second feature quantity, divides the raw data into multiple datasets using the second feature quantity, calculates the frequency distribution of the first feature quantity in each of the multiple datasets obtained from the division, and causes the processing circuit to change multiple times during the period of extracting a portion of the raw data. The following processes are repeatedly performed by setting windows: a second process, setting multiple time windows such that the sum of the periods of all time windows is shorter than the predetermined period; a third process, extracting data from the original data through the multiple time windows; a fourth process, corresponding to the division of the original data into multiple datasets, dividing the extracted data obtained by combining all the data extracted through the multiple time windows into multiple datasets, and calculating the frequency distribution of the first feature in each of the multiple datasets; and a fifth process, using the frequency distribution to determine whether the original data and the extracted data are similar, so that the processing circuit uses the extracted data similar to the original data to analyze the degree of damage to the device or the component based on the first feature and the second feature. Attached Figure Description

[0009] Figure 1 This is a schematic diagram illustrating the relationship between a data center, a vehicle, and an information processing terminal as one implementation of an information processing device.

[0010] Figure 2 This is a 3D view of the parking lock device.

[0011] Figure 3 The graphs are a portion of the raw data showing the physical quantities related to damage to the parking lock device. (a) shows the change in vehicle speed when the vehicle is in the parking position, and (b) shows the change in the vehicle's tilt angle when the vehicle is in the parking position.

[0012] Figure 4 This is a flowchart illustrating the processing flow performed by the processing circuitry of a data center.

[0013] Figure 5 It is the frequency distribution of vehicle speeds in the original data when the vehicle is in the parking position and the tilt angle is positive.

[0014] Figure 6 It is the frequency distribution of vehicle speed in the original data when the vehicle is in the parking position and the tilt angle is zero.

[0015] Figure 7 It is the frequency distribution of vehicle speed in the original data when the vehicle is in the parking position and the tilt angle is negative.

[0016] Figure 8 It concerns the corrected frequency distribution of the extracted vehicle speeds.

[0017] Figure 9 This is a cross-sectional view of the rotor.

[0018] Figure 10 The graphs are a portion of the raw data showing the physical quantities related to rotor damage, (a) showing the change in rotor angular acceleration, (b) showing the change in rotor temperature, (c) showing the change in motor coil temperature, and (d) showing the change in ATF temperature.

[0019] Figure 11 It is the frequency distribution of the rotor's angular acceleration in the original data when the rotor temperature or the motor coil temperature is lower than a predetermined temperature.

[0020] Figure 12 It is the frequency distribution of the rotor's angular acceleration in the original data when the rotor temperature or the motor coil temperature is above a predetermined temperature.

[0021] Figure 13 It relates to the corrected frequency distribution of the corrected angular acceleration in the extracted data.

[0022] Figure 14 It is the frequency distribution of the rotor's angular acceleration in the raw data when the ATF temperature is below the predetermined temperature.

[0023] Figure 15 It is the frequency distribution of the rotor's angular acceleration in the original data when the ATF temperature is above a predetermined temperature.

[0024] Figure 16 It is a cross-sectional view of the side of the differential, including the oil seal.

[0025] Figure 17 The graph shows a portion of the raw data of physical quantities related to the damage to the oil seal, (a) showing the change in the rotational speed of the drive shaft, (b) showing the change in the ATF temperature, and (c) showing the change in the external temperature.

[0026] Figure 18It is the frequency distribution of the rotational speed of the drive shaft in the original data when the ATF temperature is lower than the predetermined temperature.

[0027] Figure 19 It is the frequency distribution of the rotational speed of the drive shaft in the original data when the ATF temperature is above a predetermined temperature.

[0028] Figure 20 It is a graph showing the weighted relationship between the temperature division of ATF and the frequency of the rotational speed of the drive shaft.

[0029] Figure 21 It relates to the corrected frequency distribution of the corrected rotational speed in the extracted data.

[0030] Figure 22 It is the frequency distribution of the rotational speed of the drive shaft in the original data when the external temperature is lower than the predetermined temperature.

[0031] Figure 23 It is the frequency distribution of the rotational speed of the drive shaft in the original data when the external temperature is above a given temperature.

[0032] Figure 24 It is a cross-sectional view of the power distribution mechanism containing the planetary gear unit.

[0033] Figure 25 The graph is a portion of the raw data showing the physical quantities related to damage to the planetary gear unit. (a) shows the change in torque input to the planetary carrier, (b) shows the change in ATF temperature, (c) shows the change in the rotational speed of the second oil pump, and (d) shows the change in the vehicle's tilt angle.

[0034] Figure 26 It is the frequency distribution of the input torque to the planetary carrier in the raw data when the ATF temperature is lower than the predetermined temperature.

[0035] Figure 27 It is the frequency distribution of the input torque to the planetary carrier in the raw data when the ATF temperature is above a predetermined temperature.

[0036] Figure 28 It is a graph showing the weighted relationship between the temperature division of the ATF and the frequency of the input torque.

[0037] Figure 29 It relates to the corrected frequency distribution of the input torque extracted from the data.

[0038] Figure 30 It is the frequency distribution of the input torque to the planetary carrier in the raw data when the oil pump's rotational speed is lower than a predetermined speed.

[0039] Figure 31 It is the frequency distribution of the input torque to the planetary carrier in the raw data when the oil pump's rotational speed is above a predetermined speed.

[0040] Figure 32 It is the frequency distribution of the input torque to the planetary carrier in the raw data when the tilt angle is positive.

[0041] Figure 33 It is the frequency distribution of the input torque to the planetary carrier in the raw data when the tilt angle is zero.

[0042] Figure 34 It is the frequency distribution of the input torque to the planetary carrier in the raw data when the tilt angle is negative.

[0043] Figure 35 This is a schematic diagram showing the relationship between the drive shaft and the steering mechanism.

[0044] Figure 36 The graph shows a portion of the raw data of physical quantities related to damage to the drive shaft, (a) showing the change in torque input to the drive shaft, and (b) showing the change in steering wheel angle.

[0045] Figure 37 It is the frequency distribution of the input torque to the drive shaft in the original data when the steering wheel angle is above a predetermined rightward angle.

[0046] Figure 38 It is the frequency distribution of the input torque to the drive shaft in the raw data when the steering wheel angle is lower than a predetermined angle to the right and left.

[0047] Figure 39 It is the frequency distribution of the input torque to the drive shaft in the original data when the steering wheel angle is above a predetermined angle to the left.

[0048] Figure 40 It relates to the corrected frequency distribution of the input torque extracted from the data.

[0049] Figure 41 This is a schematic diagram of the battery and power control unit.

[0050] Figure 42 The graph is a portion of the raw data showing the physical quantities related to battery damage, (a) showing the changes in the output of the first electric generator, (b) showing the changes in the battery temperature, (c) showing the changes in the battery's SOC, (d) showing the changes in the battery's upper limit of charge, and (e) showing the changes in the battery's upper limit of discharge.

[0051] Figure 43It is the frequency distribution of the output of the first electric generator in the raw data when the battery temperature is below a predetermined temperature.

[0052] Figure 44 It is the frequency distribution of the output of the first electric generator in the raw data when the battery temperature is above a predetermined temperature.

[0053] Figure 45 It is the corrected frequency distribution of the output of the first electric generator after correction in the extracted data.

[0054] Figure 46 It is the frequency distribution of the output of the first electric generator in the raw data when the battery's SOC is lower than the first predetermined value.

[0055] Figure 47 It is the frequency distribution of the output of the first electric generator in the original data when the battery's SOC is above the first predetermined value and below the second predetermined value.

[0056] Figure 48 It is the frequency distribution of the output of the first electric generator in the original data when the battery's SOC is above the second predetermined value.

[0057] Figure 49 It is the frequency distribution of the output of the first electric generator in the raw data when the upper limit of the battery's charging power is lower than a predetermined value.

[0058] Figure 50 It is the frequency distribution of the output of the first electric generator in the original data when the upper limit of the battery charging power is above a predetermined value.

[0059] Figure 51 It is the frequency distribution of the output of the first electric generator in the raw data when the upper limit of the battery's discharge power is lower than a predetermined value.

[0060] Figure 52 It is the frequency distribution of the output of the first electric generator in the original data when the upper limit of the battery's discharge power is above a predetermined value.

[0061] Figure 53 It is equipped with Figure 1 A schematic diagram of the vehicle's cooling system, including the radiator.

[0062] Figure 54 The graph is a portion of the raw data showing the physical quantities related to radiator damage, (a) showing the change in sprung acceleration of the vehicle, (b) showing the change in coolant temperature, (c) showing the change in crankshaft speed, and (d) showing the change in water pump rotation speed.

[0063] Figure 55 It is the frequency distribution of sprung acceleration in the original data when the temperature of the cooling water is lower than the predetermined temperature.

[0064] Figure 56 It is the frequency distribution of sprung acceleration in the original data when the cooling water temperature is above a predetermined temperature.

[0065] Figure 57 It concerns the corrected frequency distribution of the corrected spring acceleration in the extracted data.

[0066] Figure 58 It is the frequency distribution of the pump rotation speed in the raw data when the cooling water temperature is lower than the predetermined temperature.

[0067] Figure 59 It is the frequency distribution of the pump rotation speed in the original data when the cooling water temperature is above a predetermined temperature.

[0068] Figure 60 It relates to the corrected frequency distribution of the pump rotation speed in the extracted data.

[0069] Figure 61 It is the frequency distribution of sprung acceleration in the original data when the crankshaft speed is outside a given range.

[0070] Figure 62 It is the frequency distribution of sprung acceleration in the original data when the crankshaft speed is within a given range.

[0071] Figure 63 It is a graph showing the relationship between the division of crankshaft speeds and the weighted acceleration on the spring.

[0072] Figure 64 It concerns the corrected frequency distribution of the corrected spring acceleration in the extracted data.

[0073] Figure 65 It is shown that it is mounted on Figure 1 A schematic diagram of the structure of a vehicle's engine.

[0074] Figure 66 The graph shows a portion of the raw data of physical quantities related to the amount of deposits in the engine's intake system, (a) showing the change in valve overlap, and (b) showing the change in road surface information segmentation.

[0075] Figure 67 This refers to the frequency distribution of valve overlap in the original data when the road surface information is divided into paved roads.

[0076] Figure 68This refers to the frequency distribution of valve overlap in the original data when the road surface information is classified as gravel road.

[0077] Figure 69 This refers to the frequency distribution of valve overlap in the original data when the road surface information is classified as dirt road.

[0078] Figure 70 It is a graph showing the relationship between valve overlap and sediment deposition.

[0079] Figure 71 It relates to the frequency distribution of sediment deposition in the extracted data. Detailed Implementation

[0080] <First Embodiment>

[0081] Below, refer to Figures 1 to 8 The first embodiment of the information processing device will be described below.

[0082] <Structure of Information Processing Systems>

[0083] Figure 1 The structure of the information processing system is shown. The information processing system includes a data center 500 with information processing devices, an information processing terminal 600, multiple vehicles 10, and a communication network 400. The data center 500 can communicate with the multiple vehicles 10 and the information processing terminal 600 via the communication network 400.

[0084] <Data Center 500 Structure>

[0085] like Figure 1 As shown, the data center 500 includes a processing circuit 510, a storage device 520, and a communication device 530. The processing circuit 510 is an information processing device, equipped with a CPU that executes processing according to a program and a ROM storing the program. The storage device 520 can store a large amount of data. The communication device 530 performs wired or wireless communication via a communication network 400. The communication device 530 includes hardware such as a network adapter, various communication software, or combinations thereof.

[0086] <Structure of Information Processing Terminal 600>

[0087] like Figure 1 As shown, the information processing terminal 600 includes a processing circuit 610, a storage device 620, and a communication device 630. The processing circuit 610 includes a CPU that executes processing according to a program and a ROM storing the program. The storage device 620 can store a large amount of data. The communication device 630 performs wired or wireless communication via a communication network 400. The communication device 630 includes hardware such as a network adapter, various communication software, or combinations thereof.

[0088] Information processing terminal 600 is, for example, a personal computer.

[0089] <Structure of Vehicle 10>

[0090] Vehicle 10 is equipped with a communication device 99. The communication device 99 transmits data acquired by vehicle 10 to data center 500 via communication network 400. Vehicle 10 includes a hybrid powertrain 20, a power control unit (hereinafter referred to as "PCU" 24), a battery 25, and a vehicle control unit 90. The vehicle control unit 90 includes a first control device 91 for controlling the hybrid powertrain 20 and a second control device 92 for controlling the PCU 24. The vehicle control unit 90 is equipped with multiple sensors for data collection. The first control device 91 has a CPU for controlling the operating state of the hybrid powertrain 20. The first control device 91 controls the hybrid powertrain 20 based on data collected by the sensors. The second control device 92 has a CPU for controlling the PCU 24. The second control device 92 controls the PCU 24 based on data collected by the sensors. Examples of data collected by the vehicle control unit 90 include crankshaft speed, electric generator speed, rotor temperature, battery SOC, and battery temperature.

[0091] The hybrid power system 20 includes an engine 21, an electric generator 23, a power distribution mechanism 60, and a drive shaft 80. The power distribution mechanism 60 includes a planetary gear unit 61, a differential 62, a reduction gear 70, an output shaft 77, and a parking lock device 30. The engine 21 transmits its output to the power distribution mechanism 60 via the engine output shaft 22. The electric generator 23 is an electric motor. Utilizing power from the battery 25 converted by the PCU 24, the electric generator 23 operates. The power distribution mechanism 60 uses multiple gears, including the planetary gear unit 61 and the reduction gear 70, to shift the outputs of the engine 21 and the electric generator 23. The power distribution mechanism 60 differentially controls the shifted power via the differential 62 and transmits it to the drive shaft 80, thereby enabling the vehicle 10 to move.

[0092] <Structure of Parking Locking Device 30>

[0093] The parking lock device 30 is housed within a housing that functions as a power distribution mechanism 60, which is also a transmission. Figure 2As shown, the parking locking device 30 includes a parking gear 31, a locking lever 32, a tapered portion 33, a rod 34, a support shaft 35, and a locking plate 36. The parking gear 31 is, for example, fixed to an output shaft 77 on which a reduction gear 70 is fixed. Because the locking lever 32 is fixed at only one end via the support shaft 35, it can rotate about the support shaft 35 as its central axis. A locking plate 36 is provided at the locking lever 32. If the locking plate 36 engages with the parking gear 31, the rotation of the parking gear 31 is mechanically restricted. Thus, the output shaft 77 is locked to prevent rotation.

[0094] Figure 2 The state of the parking lock device 30 when the parking lock is released is shown. From this state, the operation of the parking lock device 30 when the output shaft 77 is locked is explained. A rod 34 is connected to the root side of the tapered portion 33, which tapers from the root side to the front side. The tapered portion 33 is in contact with the locking rod 32. If the actuator connected to the rod 34 is activated, the rod 34 is pushed towards the front end of the tapered portion 33. At this time, the tapered portion 33 also moves simultaneously, so that the contact point between the tapered portion 33 and the locking rod 32 moves in a direction that separates from the central axis of the tapered portion 33. As a result, the locking rod 32 is pushed upward to the tapered portion 33 and rotates about the support shaft 35 in a direction closer to the parking gear 31. As a result, the locking plate 36 engages with the parking gear 31, thereby mechanically restricting the rotation of the parking gear 31.

[0095] The operation of the parking locking device 30 when the output shaft 77 is released is as follows. The actuator pulls the lever 34 back towards the root of the tapered portion 33. This causes the contact point between the tapered portion 33 and the locking lever 32 to move towards the front end of the tapered portion 33. The locking lever 32, pushed upwards by the tapered portion 33, rotates about the support shaft 35 in a direction separating it from the parking gear 31. As a result, the locking plate 36 no longer engages with the parking gear 31, and the rotation of the output shaft 77 is no longer mechanically restricted.

[0096] <Data Extraction>

[0097] Information processing terminal 600 is used to analyze the extent of damage to equipment or components mounted on a vehicle. When analyzing the extent of damage, information processing terminal 600 sends an instruction to data center 500. Upon receiving the instruction, processing circuit 510 of data center 500 uses a portion of the vast amount of data stored in storage device 520 of data center 500 for analysis. Data to be used is selected from the vast amount of data stored in storage device 520, in accordance with the purpose of the analysis. This data includes data on physical quantities related to damage to equipment or components collected using multiple sensors mounted on vehicle 10. These physical quantities are referred to as characteristic quantities. Processing circuit 510 uses the characteristic quantities to analyze the extent of damage accumulated at parking lock device 30 of a particular vehicle 10. In this case, the characteristic quantities are the vehicle speed when the gear of the vehicle 10 is in the parking position and the tilt angle of the vehicle 10.

[0098] The following describes the process by which the processing circuit 510 analyzes the extent of damage to a specific device or component of a specific vehicle 10 according to a program. The processing circuit 510 retrieves characteristic data related to the specific vehicle 10 from the storage device 520. Based on the retrieved characteristic data, it calculates the load related to the specific device or component. Based on the calculated load, the processing circuit 510 estimates the accumulated damage at the specific device or component. The processing circuit 510 sends the estimated damage result to the information processing terminal 600 for display.

[0099] To perform this analysis, the processing circuit 510 utilizes a large amount of data collected over a long period. Because the processing circuit 510 performs a massive amount of calculations, the analysis takes a considerable amount of time.

[0100] Therefore, it is considered to extract data that captures the overall characteristics of the original data from a large amount of data. If such extracted data can be extracted, the processing circuit 510 can perform parsing in a shorter time by using the extracted data. For example, in the case of predicting the damage of a component after 100,000 hours of operation, the processing circuit 510 uses extracted data equivalent to 20,000 hours extracted from the original data equivalent to 100,000 hours to predict the damage. Then, the processing circuit 510 calculates the predicted value of the damage to the equipment or component after 100,000 hours of operation by multiplying the predicted value calculated based on the extracted data equivalent to 20,000 hours by five times.

[0101] Figure 3 The raw data of the characteristic quantities related to the parking lock device 30 are shown. Figure 3 The raw data shown is a portion of the data equivalent to 100,000 hours for one vehicle. Figure 3The raw data shown includes vehicle speed and tilt angle as features.

[0102] Figure 3 (a) shows the vehicle speed in 100,000 hours of data when the vehicle is in the park position (gear 10). The speed is positive when moving forward and negative when moving backward. Figure 3 (b) shows the tilt angle of vehicle 10 in the park position from 100,000 hours of data. The tilt angle is positive when going uphill. The tilt angle is negative when going downhill.

[0103] The vehicle speed and tilt angle of vehicle 10 are correlated with the damage to the parking lock device 30 of vehicle 10. The processing circuit 510 analyzes the accumulated damage at the parking lock device 30 based on the data including vehicle speed and tilt angle as characteristic quantities.

[0104] Data extraction is achieved by cutting data from the raw data through multiple time windows. Figure 3 In the example of multiple time windows, dashed lines represent the first time window W_1, the second time window W_2, and the third time window W_3. The start and end dates of each time window are set to ensure they do not overlap. In this example, data equivalent to 20,000 hours is extracted. Therefore, the start and end dates of each time window are set such that the total length of all time windows is 20,000 hours.

[0105] Data center 500 searches for the start and end dates of each time window in the data extraction pattern used to extract features from the overall raw data. Data center 500 stores this data extraction pattern information in storage device 520. The stored data extraction pattern information is the setting of each time window found through the search.

[0106] The processing circuit 510 extracts data from the raw data based on the cutting mode information stored in the storage device 520. Then, the processing circuit 510 uses the extracted data to analyze the extent of damage to the device or component.

[0107] <Search Processing for Slicing Patterns>

[0108] Figure 4 This is a flowchart illustrating a series of processes related to the search process for the cutting mode. The processing circuit 510 of the data center 500 executes this series of processes according to the program.

[0109] like Figure 4As shown, in step S100, the processing circuit 510 obtains the raw data. The raw data is a portion of the data selected from the vast amount of data stored in the storage device 520 of the data center 500, which is matched with the purpose of parsing.

[0110] The raw data used to analyze the extent of damage to the parking lock device 30 of a vehicle 10 is the data of the vehicle 10 selected as the object from a large dataset of multiple vehicles 10.

[0111] Next, in the processing of step S110, the processing circuit 510 sets multiple time windows in order to extract data from the raw data.

[0112] exist Figure 3 In the example shown, the duration of each time window is all equal. Figure 3 As shown, the data extracted through each cutting window are data of each feature quantity within the same period.

[0113] Each time the processing circuit 510 executes step S110, it randomly sets the number of time windows, the start period of each time window, and the end period of each time window. At this time, the processing circuit 510 sets each time window in a way that prevents them from overlapping. The processing circuit 510 randomly sets multiple time windows in such a way that the period obtained by summing the periods of all time windows becomes a preset period. In the processing of step S110, the processing circuit 510, as... Figure 3 As shown, multiple time windows can also be set by fixing the duration of each time window to a constant. In the processing of step S110, the processing circuit 510 can also set multiple time windows by fixing the number of multiple time windows to a certain value.

[0114] By setting multiple time windows through step S110, the data extraction mode from the original data is determined. If the processing circuit 510 determines the extraction mode in this way, it proceeds to step S120.

[0115] In step S120, the processing circuit 510 extracts data from the original data according to the determined extraction mode. That is, in step S120, the processing circuit 510 extracts data from the original data through multiple set time windows. Then, the processing circuit 510 combines all the data extracted through the multiple time windows to form the extracted data.

[0116] Next, in step S130, the processing circuit 510 calculates the frequency distribution of the original data and the extracted data. The original data contains multiple features. One of these features is defined as the first feature, and another feature different from the first feature is defined as the second feature.

[0117] In step S130, the processing circuit 510 classifies the data containing the first feature in the original data into multiple partitions based on the data of the second feature when the first feature was collected. That is, the processing circuit 510 divides the data containing the first feature in the original data into multiple datasets based on the data of the second feature when the first feature was collected. Similarly, the processing circuit 510 classifies the data containing the first feature in the extracted data into multiple partitions based on the data of the second feature, in a manner corresponding to the partitioning of the original data. That is, the processing circuit 510, similarly to the original data, divides the data containing the first feature in the extracted data into multiple datasets based on the data of the second feature. Based on the original data and the data of the first feature in the extracted data, which are thus classified into multiple partitions, the processing circuit 510 calculates the frequency distribution of the first feature in each partition of the original data and the extracted data.

[0118] Frequency distribution categorizes data of the first feature into multiple levels, representing the distribution of the number of data points at each level. Since the total frequency of the first feature differs between the original and extracted data, a simple comparison of the frequency distributions of the original and extracted data is not possible. When extracting data equivalent to 20,000 hours from 100,000 hours of original data, the total frequency of the extracted data is approximately one-fifth of the total frequency of the original data. In this case, by multiplying the frequency of each level of the extracted data by five, a frequency distribution with the same total frequency as the original data can be obtained. Without using the above method, a comparison of the distribution of data in the original and extracted data can also be made by calculating the relative frequency distribution as the frequency distribution of the original and extracted data. The relative frequency distribution represents the percentage of the frequency of a particular level relative to the total frequency.

[0119] In the analysis of the extent of damage to the parking lock device 30, the first characteristic quantity is the vehicle speed when the vehicle 10 is in the parking position. The second characteristic quantity is the tilt angle of the vehicle 10.

[0120] Figure 5 The frequency distribution of vehicle speed in the raw data is shown when the tilt angle of vehicle 10 is positive.

[0121] Figure 6 The frequency distribution of vehicle speed in the raw data is shown when the tilt angle of vehicle 10 is zero.

[0122] Figure 7 The frequency distribution of vehicle speed in the raw data is shown when the tilt angle of vehicle 10 is negative.

[0123] like Figures 5-7As shown, in these frequency distributions, zero vehicle speed is set as the center value, and the levels are divided in such a way that the number of levels in the positive direction is equal to the number of levels in the negative direction. Figures 5-7 In the example shown, the lowest speed rating is set to "1". Figures 5-7 In the example shown, the vehicle speed is divided into 2m+1 levels, from "1" to "2m+1". The frequency distribution of the extracted data is also calculated based on the levels corresponding to the original data. Thus, the processing circuit 510 divides the vehicle speeds contained in the original and extracted data into three categories: a positive tilt angle, a negative tilt angle, and a zero tilt angle. For each of these three tilt angle categories, the processing circuit 510 calculates the frequency distribution as described above.

[0124] Next, the processing circuit 510 in Figure 4 In step S140, the error between the frequency distribution of the first feature in the original data and the frequency distribution of the first feature in the extracted data is calculated for each of the multiple partitions based on the second feature. For example, the processing circuit 510 calculates the mean absolute error (MAE). The mean absolute error (MAE) is expressed by the following formula 1.

[0125] [Formula 1]

[0126]

[0127] In Formula 1 above, "n" is the total number of ranks in the frequency distribution. For example, if it is... Figures 5-7 In the example shown, "n" is "2m+1". "i" is the number that determines the rank in the frequency distribution. For example, if it is... Figures 5-7 In the example shown, "i" is the number from "1" to "2m+1". "Y" is the frequency of the first feature at the corresponding level of the original data. "y" is the frequency of the first feature at the corresponding level of the extracted data.

[0128] As shown in Formula 1 above, for each partition, the processing circuit 510 calculates the sum of the frequency errors at each level between the frequency distribution in the original data and the first feature of the frequency distribution in the extracted data as the error.

[0129] If the error is calculated for all partitions, the processing circuit 510 proceeds to step S150. In step S150, the processing circuit 510 determines whether the calculated error for each partition is below a threshold. The threshold is a value used to determine whether extracted data with a frequency distribution close to the frequency distribution in the original data has been extracted using the set cutting pattern. This threshold is preset so that it can be determined that extracted data with a frequency distribution close to the frequency distribution in the original data has been extracted based on the error being below the threshold. This threshold can be determined to be a different value for each partition.

[0130] If, during the processing in step S150, it is determined that the error for each division is below the threshold (step S150: Yes), the processing circuit 510 records the cutting pattern. Specifically, the processing circuit 510 causes the storage device 520 to store data of the start and end times of each time window in the cutting pattern as information for determining the cutting pattern. If the cutting pattern is recorded in this way, the processing circuit 510 proceeds the processing to step S160.

[0131] On the other hand, if it is determined in step S150 that one of the errors is greater than the threshold (step S150: No), the processing circuit 510 returns the processing to step S110. That is, the processing circuit 510 starts the process of setting multiple new time windows in order to reset the time windows and extract data from the original data.

[0132] Thus, the processing circuit 510 repeatedly performs steps S110 to S150 until it can extract data similar to the original data using the frequency distribution of the first feature obtained by dividing the data by the second feature. As a result, the storage device 520 stores the cutting patterns where all errors are below the threshold.

[0133] <Calculation of fatigue damage and notification of failure prediction>

[0134] In step S160, the processing circuit 510 extracts data from the original data based on a data extraction pattern similar to the original data stored in the storage device 520.

[0135] Next, the processing circuit 510, following the program, uses the extracted data to calculate an index value representing the degree of damage to the equipment or component. For example, the index value is fatigue damage degree. The fatigue damage degree is a value from "0" to "1" that represents the proportion of damage accumulated at the equipment or component, with "1" representing the damage that causes fatigue failure at the equipment or component.

[0136] The processing circuit 510 uses the extracted data, which is part of the original data, to calculate the indicator value. Therefore, the processing circuit 510 converts the calculated indicator value to a size equivalent to the original data and calculates the corresponding indicator value. For example, if the original data is equivalent to 100,000 hours of data and the extracted data is equivalent to 20,000 hours of data, the calculated indicator value is multiplied by 5 and set as the indicator value corresponding to the original data.

[0137] Here, fatigue damage is calculated based on extracted data, serving as an indicator of the magnitude of accumulated damage at 30 points on the parking lock device.

[0138] The greater the collision energy generated between the parking gear 31 and the locking lever 32, the greater the accumulated damage due to the collision between the parking gear 31 and the locking lever 32. That is, the greater the vehicle speed when the vehicle 10 is in the parking position, the greater the accumulated damage at the parking lock device 30. When the parking lock is applied while the vehicle 10 is traveling on a slope, the greater the tilt angle, the greater the load on the parking lock device 30 caused by the vehicle weight, and the greater the accumulated damage.

[0139] As an example, the processing circuit 510 calculates the fatigue damage of the parking locking device 30 using the following method.

[0140] The processing circuit 510 calculates a frequency distribution for each segment based on data obtained by dividing the vehicle speed when the gear is in the parking position according to the tilt angle at that time. For the frequency distribution of the segment where the tilt angle of the vehicle 10 is not zero in the calculated frequency distribution, the processing circuit 510 uses Formula 2 shown below to correct the aforementioned vehicle speed.

[0141] [Formula 2]

[0142] V c =Va×sinθ

[0143] In Formula 2 above, "Vc" represents the corrected vehicle speed, "V" represents the vehicle speed in the parking position, "a" represents the coefficient, and "θ" represents the road surface gradient. θ is a positive value when the road surface gradient is uphill and a negative value when the road surface gradient is downhill.

[0144] As shown in Formula 2 above, by correcting the vehicle speed when the gear is in the parking position in accordance with the tilt angle division, the cumulative damage to the parking lock device 30 caused by the tilt angle of the vehicle 10 can be taken into account. For example, when the vehicle 10 is moving forward on an uphill slope and the gear is in the parking position, the corrected vehicle speed Vc becomes less than V. In the case of an uphill slope, a load generated by the vehicle weight is generated in the direction of reversing of the vehicle 10, so the correction is performed in a way that partially cancels out the vehicle speed in the forward direction of the vehicle 10.

[0145] Based on the corrected vehicle speed Vc obtained in this way, the processing circuit 510 gathers all the divided frequency distributions into one frequency distribution equivalent to the case where the tilt angle is zero, and calculates the new frequency distribution in the extracted data, namely the corrected frequency distribution.

[0146] Next, based on the corrected frequency distribution, the processing circuit 510 calculates the fatigue damage degree. Figure 8 The diagram illustrates an example of a corrected frequency distribution used in analyzing the extent of damage to the parking lock device 30. In this corrected frequency distribution, the corrected vehicle speed Vc is categorized into six levels, A through F. The frequency Hij for each level represents the number of data points where the corrected vehicle speed Vc is above Vi and below Vj. For example, data points where the corrected vehicle speed Vc is above V2 and below V3 are categorized into level B, and their frequency is represented as H23. An upper limit frequency Gij is determined for each level. The upper limit frequency Gij represents the maximum number of accumulated damages that would cause fatigue failure at the parking lock device 30 in cases where damage caused by the corrected vehicle speed Vc included in the corresponding level has accumulated. As an example, if G34 is L times, and L collisions occur between the parking gear 31 and the locking lever 32 within the range where the corrected vehicle speed Vc is above V3 and below V4, then fatigue failure will occur at the parking lock device 30. The processing circuit 510 uses the frequency Hij of each level in the corrected frequency distribution and the upper limit frequency Gij of each level to calculate the fatigue damage degree according to the following formula 3.

[0147] [Formula 3]

[0148] Fatigue damage degree = H 12 / G 12 +H 23 / G 23 +…+H 56 / G 56 +H 67 / G 67

[0149] If the processing circuit 510 calculates the fatigue damage degree based on the above formula 3, then the processing is advanced to... Figure 4 The step S170 shown.

[0150] In step S170, the processing circuit 510 determines whether the fatigue damage degree is above a boundary value. The boundary value is a value used to predict that the likelihood of fatigue failure is higher when the fatigue damage degree is above a boundary value. For example, a boundary value of "0.9" can be set. In this case, based on the condition that 90% of the fatigue level leading to fatigue failure has been reached, the likelihood of fatigue failure can be predicted to be higher.

[0151] If, during the processing in step S170, it is determined that the fatigue damage degree exceeds the boundary value (step S170: Yes), the processing circuit 510 proceeds to step S180. In step S180, the processing circuit 510 outputs the fatigue damage degree and fault prediction. Specifically, the processing circuit 510 sends the fatigue damage degree and fault prediction to the information processing terminal 600 for display.

[0152] Fault prediction, for example, indicates a prediction of fatigue failure. Thus, the processing circuit 510 notifies the processor that fatigue failure has been predicted if the calculated fatigue damage exceeds a threshold value.

[0153] If, during the processing in step S170, it is determined that the fatigue damage degree is below the boundary value (step S170: No), the processing circuit 510 proceeds to step S190. In step S190, the processing circuit 510 outputs the fatigue damage degree. Specifically, the processing circuit 510 of the data center 500 sends the calculated fatigue damage degree to the information processing terminal 600 for display.

[0154] If step S180 or step S190 is performed, the processing circuit 510 terminates the above series of program-based processes.

[0155] <Function of the first embodiment>

[0156] The data center 500, which is the information processing device in this embodiment, extracts a portion of the raw data collected within a predetermined period using multiple sensors mounted on the vehicle 10, and analyzes the extent of damage accumulated at the parking lock device 30.

[0157] Data center 500 includes a processing circuit 510 that performs processing according to a program. The raw data includes the vehicle speed of vehicle 10 when it is in a parked position as a first feature. The raw data also includes data on the tilt angle of vehicle 10 as a second feature. In this data center 500, the processing circuit 510 performs a search process. This process includes a first process (step S130), which divides the data of the first feature into multiple parts using the second feature included in the raw data, and calculates the frequency distribution of the first feature in the raw data for each part. The search process includes a second process (step S110), which sets multiple time windows to extract data from a portion of the raw data, such that the sum of the periods of all time windows is shorter than the overall period of the raw data. The search process includes a third process (step S120) to extract data from the raw data through the multiple time windows. The data obtained by combining all the data extracted through the multiple time windows is the extracted data. The search process includes a fourth process (step S130), which divides the data of the first feature quantity into multiple partitions corresponding to the second feature quantity of the original data, and calculates the frequency distribution of the extracted data with respect to the first feature quantity for each partition. The search process includes a fifth process (steps S140 and S150) which calculates the error between the frequency distribution of the original data and the frequency distribution of the extracted data to determine whether the original data and the extracted data are similar. After executing the first process, the processing circuit 510 performs a search process that repeatedly executes the second to fifth processes by changing the settings of multiple time windows. Then, the processing circuit 510 extracts the extracted data that meets the condition that the error is below a threshold. The processing circuit 510 uses the extracted data with the error below the threshold as an indicator value of damage to calculate the fatigue damage degree (step S160).

[0158] According to the data center 500, the distribution of characteristic quantities related to damage to the parking lock device 30 can be analyzed using extracted data similar to the original data. Therefore, the data center 500 can obtain analysis results that are close to those obtained using the original data for damage analysis.

[0159] The extracted data obtained through Data Center 500 is obtained by cutting a portion of the original data. Therefore, the amount of extracted data is smaller compared to the original data. The more data used in parsing, the longer the processing time required for parsing the degree of corruption. By using extracted data, Data Center 500 is able to shorten the parsing time compared to using the original data.

[0160] <Effects of the first embodiment>

[0161] (1-1) According to the data center 500 of the information processing apparatus as described in the first embodiment, data suitable for analyzing the degree of damage to equipment or components mounted on a vehicle can be extracted from raw data. Therefore, according to the above-described information processing apparatus, the degree of damage to equipment or components can be analyzed in a shorter time compared to the case of using raw data.

[0162] (1-2) According to the information processing method of the first embodiment, data suitable for analyzing the degree of damage to equipment or components mounted on a vehicle can be extracted from raw data. Therefore, according to the above information processing method, the degree of damage to equipment or components can be analyzed in a shorter time compared to the case of using raw data.

[0163] (1-3) The program included in the processing circuit 510 of the first embodiment enables the processing circuit 510 to extract data suitable for analyzing the degree of damage to the equipment or components mounted on the vehicle from the raw data. Therefore, according to the above program, the processing circuit 510 can analyze the degree of damage to the equipment or components in a shorter time compared to the case of using the raw data.

[0164] (1-4) In the fifth process (steps S140 and S150), the processing circuit 510 of the data center 500 calculates the error between the frequency distribution of the original data and the frequency distribution of the extracted data for each data obtained by partitioning. When the calculated error of each partition is below the threshold, it is determined that the original data and the extracted data are similar.

[0165] Based on the aforementioned data center 500, the extracted data with the smallest error in any partition is used to parse the corrupted data. Therefore, based on the aforementioned data center 500, it is possible to appropriately determine whether the original data and the extracted data are similar, regardless of the difference in the total frequency of each partition.

[0166] (1-5) The processing circuit 510 of the data center 500 corrects the first feature contained in the extracted data according to the division, and analyzes the degree of damage to the equipment or component based on the corrected first feature.

[0167] Even data with the same first feature quantity will have different impacts on the analysis results if the second feature quantity is different. By correcting the first feature quantity according to the partitioning, the influence of the difference in the second feature quantity can be incorporated into the corrected first feature quantity. Therefore, according to the aforementioned data center 500, analysis that also reflects the influence of the second feature quantity can be performed based on the corrected first feature quantity.

[0168] (1-6) The data center 500 analyzes the extent of damage to the parking lock device 30 that prevents the rotation of the output shaft 77 within the power distribution mechanism 60, which functions as a transmission. The processing circuit 510 of the data center 500 sets the vehicle speed when the transmission is in the parking position as the first characteristic quantity and the vehicle tilt angle as the second characteristic quantity.

[0169] The parking lock device 30 suffers cumulative damage due to collisions between the parking gear 31 and the locking lever 32. The more collisions occur, the more severe the cumulative damage. The greater the collision energy generated between the parking gear 31 and the locking lever 32, the greater the cumulative damage. In other words, the higher the vehicle speed when the gear is in the parking position, the greater the cumulative damage at the parking lock device 30. When the vehicle 10 is traveling on a slope and the gear is in the parking position, the magnitude of the cumulative damage at the parking lock device 30 also varies depending on the angle of inclination. The larger the angle of inclination, the greater the load exerted on the parking lock device 30 by the vehicle's weight, thus resulting in greater cumulative damage at the parking lock device 30. The data center 500 uses the two physical quantities that affect the magnitude of the cumulative damage at the parking lock device 30 as feature quantities to obtain extracted data.

[0170] Therefore, based on the aforementioned data center 500, data suitable for analyzing the extent of damage to the parking lock device 30 can be extracted.

[0171] <Second Implementation>

[0172] Next, refer to Figures 9-15 This section describes a second embodiment of the information processing device. Furthermore, the second embodiment is an information processing device for analyzing the degree of damage to the rotor 40 of an electric generator 23, which is one of the rotating machines mounted on a vehicle. The second embodiment differs from the first embodiment in that the device or component used to analyze the degree of damage is different. In the following description, the parts that differ from the first embodiment will be mainly explained. Detailed descriptions of components that are repeated in the first embodiment are omitted. In the second embodiment, the information processing device for analyzing the degree of damage is also the processing circuit 510 of the data center 500.

[0173] <Structure of Electric Generator 23>

[0174] For reference Figure 1 As can be explained, the hybrid power system 20 installed in the vehicle 10 includes an electric generator 23. The electric generator 23 includes a rotor 40 and a stator.

[0175] like Figure 9 As shown, the rotor 40 includes a rotor core 41, a rotor shaft 44, an end plate 42, and an end plate 43. Figure 9As shown, the rotor shaft 44 includes a central shaft 45 and a mounting portion 46. The mounting portion 46 is directly connected to the rotor core 41 and end plates 42 and 43. At one end of the mounting portion 46, a small flange 47 and a large flange 48 are provided. At the other end of the mounting portion 46, a riveting portion 49 is provided.

[0176] The rotor core 41 is formed by stacking annular electromagnetic steel plates along the rotation axis of the rotor shaft 44. At both ends of the stacked rotor core 41, annular end plates 42 and 43 are disposed.

[0177] like Figure 9 As shown, the end plate 42 at the left end of the stacked rotor core 41 is connected to the large flange 48 of the rotor shaft 44 on the side opposite to the surface that contacts the stacked rotor core 41. Regarding the end plate 42, a protrusion on the inner circumferential surface of the annulus engages with a recess on the outer circumferential surface of the small flange 47 on the rotor shaft 44. Regarding the stacked rotor core 41, a protrusion on the inner circumferential surface of the annulus of the rotor core 41 engages with a recess in the mounting portion 46 on the rotor shaft 44. By engaging the recess and protrusion in this way, the rotor core 41 and the end plate 42 can rotate integrally with the rotor shaft 44. The end plate 43 at the right end of the stacked rotor core 41 is fixed to the rotor shaft 44 by riveting the end plate 43 radially outward while being pressed towards the rotor core 41 via a riveting portion 49.

[0178] The rotor 40 rotates around the central axis 45 of the rotor shaft 44. The rotational speed of the rotor 40 changes frequently to match the acceleration and deceleration of the vehicle. If a sudden change in rotational speed occurs, damage accumulates at the rotor 40 due to inertia. The end plate 43 is fixed to the rotor shaft 44 while being pressed against the rotor core 41 by the riveting part 49. In this structure, if a sudden change in rotational speed occurs, a slight misalignment occurs between the rotor core 41 and the end plate 43, causing the rotor core 41 to collide with the end plate 43, resulting in accumulated damage. As a result, loosening occurs at the riveting part 49 between the rotor shaft 44 and the end plate 43, leading to rotor 40 failure.

[0179] <Data Extraction>

[0180] The information processing terminal 600 sends an indication to the data center 500 regarding the extent of damage to the rotor 40, which is a rotating machine. Then, similarly to the first embodiment, the data center 500 performs data extraction from the raw data through multiple time windows.

[0181] Figure 10 The original data of characteristic quantities related to the rotor 40 of a particular vehicle 10 are shown. Figure 10 The raw data shown is a portion of the data equivalent to 100,000 hours from vehicle 10, which is the object of analysis. Figure 10 The raw data shown includes the angular acceleration of rotor 40, the temperature of rotor 40, the temperature of the motor coil of electric generator 23 (which is related to the temperature of rotor 40), and the temperature of the refrigerant cooling rotor 40 (ATF - automatic transmission fluid) as characteristic quantities. These characteristic quantities are physical quantities that are related to damage to rotor 40. Figure 10 (a) shows the angular acceleration of rotor 40, i.e., the change in rotational speed. Figure 10 (b) shows the temperature of rotor 40. Figure 10 (c) shows the temperature of the motor coil of the electric generator 23. Figure 10 (d) shows the ATF temperature. Data center 500 identifies a cutting pattern for extracting features from the raw data containing the aforementioned characteristic quantities. Using the extracted data extracted based on the cutting pattern identified by data center 500, processing circuit 510 analyzes the accumulated damage at the rotor 40 of vehicle 10, which is the object of analysis.

[0182] <Search Processing for Slicing Patterns>

[0183] like Figure 4 As shown, the processing circuit 510 performs the same series of processes as in the first embodiment according to the program.

[0184] In step S100, the processing circuit 510 acquires raw data for a specific vehicle 10. The raw data includes data used to analyze the extent of damage to the rotor 40 of the vehicle 10, which is the subject of the analysis.

[0185] Next, in the processing of step S110, the processing circuit 510 targets... Figure 10 The raw data of characteristic quantities related to damage to rotor 40 shown are, in the same manner as in the first embodiment, determined by setting multiple time windows to determine the cutting mode. The multiple time windows are set in such a way that the sum of the periods of all the time windows is shorter than the total period of the raw data.

[0186] In step S120, the processing circuit 510, similarly to the first embodiment, extracts data by cutting data through multiple time windows based on the cutting mode determined in step S110.

[0187] In step S130, the processing circuit 510 calculates the frequency distribution of characteristic quantities related to the damage to the rotor 40. In analyzing the degree of damage to the rotor 40, the first characteristic quantity is the angular acceleration of the rotor 40. The second characteristic quantity is the temperature of the rotor 40 or the temperature of the motor coil of the electric generator 23. The processing circuit 510 divides the angular acceleration of the rotor 40 into multiple values ​​according to the temperature of the rotor 40 or the temperature of the motor coil, and similarly to the first embodiment, calculates the frequency distribution of the original data and the extracted data obtained by combining all data extracted through multiple time windows. Figure 11 The frequency distribution of the angular acceleration of the rotor 40 is shown in the raw data when the temperature of the rotor 40 or motor coil of the vehicle 10, which is the object of analysis, is below a predetermined temperature. Figure 12 This shows the frequency distribution of the angular acceleration of the rotor 40 in the raw data when the temperature of the rotor 40 or motor coil of the vehicle 10, which is the object of analysis, is above a predetermined temperature. For example... Figure 11 and Figure 12 As shown, in these frequency distributions, angular acceleration is set to zero as the minimum level, and angular acceleration is divided into m levels from "1" to "m". The frequency distribution of the extracted data is also calculated based on the levels corresponding to the original data. In this example, the processing circuit 510 divides the temperature of the rotor 40 or the temperature of the motor coil, which is included in the original data and the extracted data, into two categories: a category below a predetermined temperature and a category above a predetermined temperature. For each of the two categories of rotor 40 temperature or motor coil temperature, the processing circuit 510 calculates the frequency distribution as described above.

[0188] Next, the processing circuit 510 in Figure 4 In the processing of step S140 shown, similarly to the first embodiment, for each of the multiple partitions based on the second feature, the error between the frequency distribution of the first feature in the original data and the frequency distribution of the first feature in the extracted data is calculated. Regarding the error, for example, similar to the first embodiment, the formula for calculating the mean absolute error (MAE), i.e., Formula 1, can be used. In this case, if... Figure 11 and Figure 12 In the example shown, "n" is "m". Similarly, "i" is the number from "1" to "m". If the error is calculated with respect to the entire division using Formula 1 above, the processing circuit 510 advances the processing to step S150.

[0189] The processing in step S150 is the same as that performed in the first embodiment. The processing circuit 510 uses the error of each segment to determine whether the original data and the extracted data are similar. Then, the processing circuit 510 repeatedly performs steps S110 to S150 by changing the settings of multiple time windows until it can extract data with an error below a threshold. As a result, the storage device 520 stores the cutting patterns where each error is below the threshold. In this way, the processing circuit 510 obtains a cutting pattern for extracted data similar to the original data.

[0190] <Calculation of fatigue damage and notification of failure prediction>

[0191] In step S160, the processing circuit 510 extracts data from the original data based on a data extraction pattern similar to the original data stored in the storage device 520 after processing prior to step S150.

[0192] In the second embodiment, similar to the first embodiment, fatigue damage is calculated based on the extracted data, serving as an index value representing the degree of damage accumulated at rotor 40.

[0193] If the rotational speed of rotor 40 changes, the end plate 43 constituting rotor 40 will loosen due to the inertia acting on rotor 40, resulting in repeated collisions between end plate 43 and rotor core 41 and cumulative damage. The greater the change in rotational speed, i.e., the greater the angular acceleration, the greater the cumulative damage. The strength of end plate 43 and rotor core 41 changes with temperature, so the magnitude of the damage accumulated due to collisions varies depending on the temperature of rotor 40 at the time of the collision.

[0194] As an example, the processing circuit 510 calculates the fatigue damage degree of the rotor 40 using the following method.

[0195] The processing circuit 510 calculates the frequency distribution for each division of the angular acceleration of the rotor 40 based on the current temperature of the rotor 40 or the temperature of the motor coil. One of the calculated divisions is designated as the reference division. Next, for the data in divisions other than the reference division, the angular acceleration data included in those divisions is corrected according to the temperature division. Regarding this correction, the thermal expansion coefficients and thermal conductivity of the components constituting the rotor 40 can be considered, and any method that reflects the changes in the riveting strength of the riveting part 49 and the pressure input to the rotor core 41 and the end plate 43 caused by temperature can be used. As an example, if the corrected angular acceleration ωc is calculated using a formula incorporating the temperature of the rotor 40 or the temperature of the motor coil, the subsequent processing is as follows.

[0196] Based on the corrected angular acceleration ωc obtained using the formula described above, the processing circuit 510 aggregates all the divided frequency distributions into a single frequency distribution determined as a reference division, and calculates a new frequency distribution in the extracted data, namely the corrected frequency distribution. Next, the processing circuit 510 calculates the fatigue damage degree based on the corrected frequency distribution. Figure 13 The diagram shows an example of a corrected frequency distribution in the analysis of the degree of damage to rotor 40. In this corrected frequency distribution, the corrected angular acceleration ωc is classified into six levels, A to F. The frequency Hij for each level represents the number of data where the corrected angular acceleration ωc is ωi or higher and lower than ωj. For example, data where the corrected angular acceleration ωc is ω2 or higher and lower than ω3 are classified into level B, and their frequency is represented as H23. An upper limit frequency Gij is determined for each level. The upper limit frequency Gij represents the maximum number of damage accumulations that can lead to fatigue failure at rotor 40 when damage caused by the corrected angular acceleration ωc included in the corresponding level is accumulated. As an example, if G34 is L times, fatigue failure will occur at rotor 40 if L times of rotational speed changes occur within the range where the corrected angular acceleration ωc is ω3 or higher and lower than ω4. Processing circuit 510 uses the frequency Hij for each level and the upper limit frequency Gij for each level in the corrected frequency distribution to calculate the fatigue damage degree according to Formula 3, similar to the first embodiment. If the processing circuit 510 calculates the fatigue damage degree, it advances the processing to the next step. Figure 4 The step S170 shown.

[0197] Regarding the subsequent steps S170 to S190, the processing circuit 510 performs the same processing as in the first embodiment.

[0198] <Function of the Second Embodiment>

[0199] The data center 500 of the information processing device in the second embodiment extracts a portion of the raw data collected over a predetermined period using multiple sensors mounted on the vehicle 10, and analyzes the extent of damage accumulated at the rotor 40 of the electric generator 23.

[0200] Data center 500 includes processing circuit 510 that performs processing according to a program. The raw data includes the angular acceleration of the rotor 40, which is a rotating machine mounted on vehicle 10, as a first characteristic quantity. The raw data includes the temperature of the rotor 40, or the temperature of the motor coil of the electric generator 23, which is related to the temperature of the rotor 40, as a second characteristic quantity. In this data center 500, processing circuit 510 performs a search process. The search process includes a first process (step S130), in which the data of the first characteristic quantity is divided into multiple parts according to the second characteristic quantity included in the raw data, and the frequency distribution of the raw data with respect to the first characteristic quantity is calculated for each part. The search process includes a second process (step S110), in which multiple time windows are set to extract data from a portion of the raw data, such that the sum of the periods of all time windows is shorter than the overall period of the raw data. The search process includes a third process (step S120) to extract data from the raw data through multiple time windows. The data obtained by combining all the data extracted through the multiple time windows is the extracted data. The search process includes a fourth process (step S130), which divides the data of the first feature quantity into multiple partitions corresponding to the second feature quantity of the original data, and calculates the frequency distribution of the extracted data with respect to the first feature quantity for each partition. The search process includes a fifth process (steps S140 and S150) which calculates the error between the frequency distribution of the original data and the frequency distribution of the extracted data to determine whether the original data and the extracted data are similar. After executing the first process, the processing circuit 510 performs a search process that repeatedly executes the second to fifth processes by changing the settings of multiple time windows. Then, the processing circuit 510 extracts the extracted data with an error below a threshold. Using the extracted data with an error below the threshold, the processing circuit 510 calculates the fatigue damage degree as an indicator value of damage (step S160).

[0201] According to the data center 500, the distribution of characteristic quantities related to the damage to the rotor 40 can be analyzed using extracted data similar to the original data. Therefore, the data center 500 can obtain analysis results that are close to those obtained using the original data for damage analysis.

[0202] The extracted data obtained through Data Center 500 is obtained by cutting a portion of the original data. Therefore, the amount of extracted data is smaller compared to the original data. The more data used in parsing, the longer the processing time required for parsing the degree of corruption. By using extracted data, Data Center 500 is able to shorten the parsing time compared to using the original data.

[0203] <Effects of the Second Embodiment>

[0204] In addition to the effects of the first embodiment (1-1) to (1-5), the second embodiment also has the following effects.

[0205] (2-1) Data center 500 analyzes the degree of damage to the rotor 40 of the electric generator 23, which is a rotating machine. The processing circuit 510 of data center 500 sets the angular acceleration of rotor 40 as the first characteristic quantity and the temperature of rotor 40 or the temperature of motor coil of electric generator 23 as the second characteristic quantity.

[0206] If the rotational speed of rotor 40 changes, the end plate 43 constituting rotor 40 will loosen due to the inertia acting on rotor 40, resulting in repeated collisions between end plate 43 and rotor core 41. This causes wear and cumulative damage to rotor core 41. The greater the change in rotational speed, i.e., the greater the angular acceleration, the greater the cumulative damage. The strength of the components constituting rotor 40 changes with temperature; therefore, the magnitude of damage accumulated due to collisions varies depending on the temperature of rotor 40 at the time of the collision. The data center 500 uses the two physical quantities that affect the magnitude of accumulated damage at rotor 40 as feature quantities to obtain extracted data. Therefore, based on the data center 500, data suitable for analyzing the degree of damage to rotor 40 can be extracted.

[0207] <Amendment to the Second Embodiment>

[0208] The second embodiment described above can be implemented by modifications as described below. The second embodiment and the following modifications to the second embodiment can be combined with each other to implement them without creating technical inconsistencies.

[0209] The aforementioned data center 500 divides the rotational speed of rotor 40 into multiple values ​​based on the temperature of rotor 40 or the temperature of the motor coil of electric generator 23 for data extraction. The data center 500 can use the temperature of the refrigerant cooling rotor 40 as a second characteristic quantity instead of the temperature of rotor 40 or the temperature of the motor coil of electric generator 23. The temperature of the refrigerant cooling rotor 40 is, for example, the ATF temperature. The processing circuit 510 of the data center 500 divides the angular acceleration of rotor 40 into multiple values ​​based on the ATF temperature and calculates the frequency distribution of the raw data and the extracted data. For example, Figure 14 The frequency distribution of the angular acceleration of the rotor 40 is shown in the raw data when the ATF temperature of the vehicle 10, which is the subject of analysis, is below a predetermined temperature. Figure 15 The frequency distribution of the rotor 40's angular acceleration in the raw data when the ATF temperature of the vehicle 10 being analyzed is above a predetermined temperature is shown. In these frequency distributions, zero angular acceleration is set as the minimum level, and the angular acceleration is divided into m levels from "1" to "m". The processing circuit 510 calculates for each division of the ATF temperature... Figure 14 and Figure 15 The frequency distribution of the aforementioned angular acceleration in the original data and extracted data is shown. Data Center 500 can extract the extracted data from the original data using the frequency distribution calculated according to ATF temperature.

[0210] The strength of the components constituting rotor 40 changes with temperature; therefore, the extent of damage accumulated due to the impact varies depending on the temperature of rotor 40 at the time of the impact. The temperature of the refrigerant cooling rotor 40, i.e., the ATF temperature, affects the temperature of rotor 40. Data center 500 uses the ATF temperature, which affects the extent of damage accumulated at rotor 40, as a second characteristic quantity to obtain extracted data. Therefore, according to this data center 500, data suitable for analyzing the degree of damage to rotor 40 can be extracted.

[0211] The data center 500 described above analyzes the extent of damage to the rotor 40 of the electric generator 23 mounted on the hybrid power mechanism 20 as a rotating machine. However, the rotating machine whose extent of damage can be analyzed by the data center 500 is not limited to the rotor 40 of the electric generator 23 described above. Similar to the rotor 40, it can be used to analyze the extent of damage to other rotating machines composed of combinations of multiple components. As an example, it can be used to analyze the extent of damage to crankshafts, geared shafts, and other rotating equipment or components.

[0212] <Third Implementation>

[0213] Next, refer to Figures 16-23 This section describes a third embodiment of the information processing device. Furthermore, the third embodiment is an information processing device that analyzes the degree of damage to an oil seal, which is one of the sealing members that slides in contact with a rotating body mounted on a vehicle. The third embodiment differs from the first embodiment in that the device or component that analyzes the degree of damage is different. In the following description, the parts that differ from the first embodiment will be mainly explained. Detailed descriptions of components that are repeated in the first embodiment are omitted. In the third embodiment, the information processing device that analyzes the degree of damage is also the processing circuit 510 of the data center 500.

[0214] <Structure of differential side 50>

[0215] Figure 16 This is a schematic cross-sectional view of the differential side 50. The differential side 50 is the portion of the drive shaft 57 that protrudes from the housing 56 housing the differential assembly 62 towards the outside of the housing 56. Figure 16 In the middle, the right side of the figure is the space inside the outer shell 56.

[0216] like Figure 16As shown, the differential side 50, in addition to the housing 56 and drive shaft 57, also includes an oil seal 51. The oil seal 51 has a core 52, a main lip 53, a secondary lip 54, and a side protrusion 55. Figure 16 The cross-sectional shape of the component shown is an annular. The oil seal 51 is fitted and fixed to the opening 58 of the housing 56. The drive shaft 57 is a structure that passes through the opening 58 of the housing 56 and the inner side of the annulus of the oil seal 51 and protrudes outward from the housing 56.

[0217] The main lip 53 and secondary lip 54 of the oil seal 51 slide in contact with the drive shaft 57. Therefore, the main lip 53 and secondary lip 54 wear due to friction with the drive shaft 57, leading to accelerated deterioration. The oil seal 51 is a component that simultaneously prevents lubricating oil from leaking out of the housing and prevents dust from entering from the outside. That is, the oil seal 51 is in contact with both the lubricating oil inside the housing and the outside air. Therefore, the deterioration of the oil seal 51 is affected by the temperature of the lubricating oil and the outside air temperature.

[0218] <Data Extraction>

[0219] The information processing terminal 600 sends an indication to the data center 500 regarding the degree of damage to the sealing member, i.e., the oil seal 51, which is in sliding contact with the drive shaft 57, which is a rotating body. Thus, similar to the first embodiment, the data center 500 performs data processing by extracting data from the raw data through multiple time windows.

[0220] Figure 17 The raw data of characteristic quantities related to the oil seal 51 of a specific vehicle 10 are shown. Figure 17 The raw data shown is a portion of the data equivalent to 100,000 hours from vehicle 10, which is the object of analysis. Figure 17 The raw data shown includes the rotational speed of drive shaft 57, the temperature of the fluid sealed by oil seal 51 (ATF), and the external air temperature as characteristic quantities. These characteristic quantities are physical quantities that are related to damage to oil seal 51. Figure 17 (a) shows the rotational speed of drive shaft 57. Figure 17 (b) shows the temperature of the ATF. Figure 17 (c) shows the external temperature. Data center 500 identifies a cutting pattern for extracting features from the raw data containing the aforementioned characteristic quantities. Using the extracted data extracted based on the cutting pattern identified by data center 500, processing circuit 510 analyzes the accumulated damage at the oil seal 51 of the vehicle 10, which is the object of analysis.

[0221] <Search Processing for Slicing Patterns>

[0222] like Figure 4As shown, the processing circuit 510 performs the same series of processes as in the first embodiment according to the program.

[0223] In step S100, the processing circuit 510 acquires raw data for a specific vehicle 10. The raw data includes data used to analyze the extent of damage to the oil seal 51 of the vehicle 10 being analyzed.

[0224] Next, in the processing of step S110, the processing circuit 510 targets... Figure 17 The raw data of the characteristic quantities related to the damage to oil seal 51 shown are, in the same manner as in the first embodiment, determined by setting multiple time windows to determine the cutting mode. The multiple time windows are set in such a way that the sum of the periods of all the time windows is shorter than the period of the entire raw data.

[0225] In step S120, the processing circuit 510, similarly to the first embodiment, extracts data by cutting data through multiple time windows based on the cutting mode determined in step S110.

[0226] In step S130, the processing circuit 510 calculates the frequency distribution of characteristic quantities related to the damage to the oil seal 51. In analyzing the degree of damage to the oil seal 51, the first characteristic quantity is the rotational speed of the drive shaft 57. The second characteristic quantity is the ATF temperature. The processing circuit 510 divides the rotational speed of the drive shaft 57 into multiple values ​​according to the ATF temperature, and, similarly to the first embodiment, calculates the frequency distribution of the original data and the extracted data obtained by combining all data extracted through multiple time windows. Figure 18 The frequency distribution of the rotational speed of the drive shaft 57 is shown in the raw data when the ATF temperature of the vehicle 10, which is the subject of the analysis, is lower than a predetermined temperature. Figure 19 The frequency distribution of the rotational speed of the drive shaft 57 in the raw data is shown when the ATF temperature of the vehicle 10, the object of analysis, is above a predetermined temperature. For example... Figure 18 and Figure 19 As shown, in these frequency distributions, zero rotational speed is set as the smallest level, and the rotational speed is divided into m levels from "1" to "m". The frequency distribution of the extracted data is also calculated based on the levels corresponding to the original data. In this example, the processing circuit 510 divides the ATF temperature into two categories based on the rotational speed of the drive shaft 57 included in the original data and the extracted data: a category below a predetermined temperature and a category above a predetermined temperature. The processing circuit 510 calculates the frequency distribution as described above for each of the two ATF temperature categories.

[0227] Next, the processing circuit 510 in Figure 4In the processing of step S140 shown, similarly to the first embodiment, for each of the multiple partitions based on the second feature, the error between the frequency distribution of the first feature in the original data and the frequency distribution of the first feature in the extracted data is calculated. Regarding the error, for example, similar to the first embodiment, the formula for calculating the mean absolute error (MAE), i.e., Formula 1, can be used. In this case, if... Figure 18 and Figure 19 In the example shown, "n" is "m". Similarly, "i" is the number from "1" to "m". If the error is calculated with respect to the entire division using Formula 1 above, then the processing circuit 510 advances the processing to step S150.

[0228] The processing in step S150 is the same as that performed in the first embodiment. The processing circuit 510 uses the error of each segment to determine whether the original data and the extracted data are similar. Then, the processing circuit 510 repeatedly performs steps S110 to S150 by changing the settings of multiple time windows until it can extract data with an error below a threshold. As a result, the storage device 520 stores the cutting patterns where each error is below the threshold. In this way, the processing circuit 510 obtains a cutting pattern for extracted data similar to the original data.

[0229] <Calculation of Degradation and Prediction of Failure>

[0230] In step S160, the processing circuit 510 extracts data from the original data based on a data extraction pattern similar to the original data stored in the storage device 520 after processing prior to step S150.

[0231] In the third embodiment, the degree of degradation is calculated based on the extracted data, serving as an indicator of the extent of damage accumulated at the oil seal 51. Regarding the degree of degradation, for example, the degree of degradation leading to undesirable conditions related to the oil seal 51 can be set to "1," and can be defined as a value from "0" to "1" representing the proportion of degradation accumulated at the oil seal 51. Degradation of the oil seal refers, for example, to wear or deformation of the oil seal. Such degradation may lead to undesirable conditions such as lubricating oil leakage or insufficient lubrication of the differential device.

[0232] The oil seal 51, which slides in contact with the drive shaft 57, wears and deteriorates due to friction with the drive shaft 57. Therefore, the higher the rotational speed of the drive shaft 57, the more easily the deterioration caused by sliding contact is aggravated. Regarding the oil seal 51, the higher the temperature of the oil seal 51, the more easily the deterioration is aggravated. The temperature of the oil seal 51 is affected by the temperature of the ATF sealed by the oil seal 51.

[0233] As an example, the processing circuit 510 calculates the degree of deterioration of the oil seal 51 using the following method.

[0234] The processing circuit 510 calculates a frequency distribution for each division based on data obtained by dividing the rotational speed of the drive shaft 57 according to the current ATF temperature. The division with the lowest ATF temperature among the multiple divisions for which the frequency distribution is calculated is determined as the reference division. Next, the processing circuit 510 performs a process to correct the frequency distribution of divisions other than the division with the lowest ATF temperature to a frequency distribution comparable to that of the division with the lowest ATF temperature. Figure 20 This is a graph showing the relationship between the temperature divisions of the ATF and the weighted average of the frequencies of rotational speed. (Using...) Figure 20 This allows the frequency distribution of a certain partition to be corrected to a frequency distribution comparable to that of the partition with the lowest ATF temperature. Figure 20 In this context, the lowest temperature is designated as 'a', and the ATF temperatures are divided into five categories: a through e. For example, when correcting the frequency distribution in temperature category c to a distribution equivalent to that in temperature category a, the frequencies Fi of each level of the corrected frequency distribution are used... Figure 20 Use Formula 4 below to calculate.

[0235] [Formula 4]

[0236]

[0237] In Formula 4 above, "i" is the number that determines the rank in the frequency distribution. If it is... Figure 18 and Figure 19 In the example shown, "i" is the number from "1" to "m". "fi" is the frequency of level i in temperature division c. "Fi" is the frequency of level i when fi is corrected to the frequency in temperature division a. "C3" and "C1" are as follows: Figure 20 The figures shown are the coefficients used in the weighting of temperature division c and temperature division a, respectively.

[0238] Processing circuit 510 uses Formula 4 above to calculate the frequency of each level of the corrected frequency distribution. Then, processing circuit 510 adds the calculated frequency of each level to the frequency of each level corresponding to the frequency distribution with the lowest ATF temperature. In this way, processing circuit 510 merges multiple frequency distributions into the frequency distribution with the lowest ATF temperature, and calculates the new frequency distribution, i.e., the corrected frequency distribution, from the extracted data.

[0239] Next, the processing circuit 510 calculates the degree of degradation based on the corrected frequency distribution. Figure 21The diagram shows an example of a corrected frequency distribution in the analysis of the degree of damage to oil seal 51. In this corrected frequency distribution, the corrected rotational speed Rc is classified into six levels, A to F. The frequency Hij for each level represents the number of data where the corrected rotational speed Rc is above Ri and below Rj. For example, data where the corrected rotational speed Rc is above R2 and below R3 is classified into level B, and its frequency is represented as H23. An upper limit frequency Gij is determined for each level. The upper limit frequency Gij represents the maximum number of times damage accumulates at oil seal 51 when damage caused by the corrected rotational speed Rc included in the corresponding level accumulates. As an example, if G34 is L times, and L times of rotation of drive shaft 57 occurs within the range where the corrected rotational speed Rc is above R3 and below R4, then a poor condition occurs at oil seal 51. The processing circuit 510 uses the frequency Hij of each level in the corrected frequency distribution and the upper limit frequency Gij of each level to calculate the degree of degradation according to the following formula 5.

[0240] [Formula 5]

[0241] Deterioration degree = H 12 / G 12 +H 23 / G 23 +…+H 56 / G 56 +H 67 / G 67

[0242] If the processing circuit 510 calculates the degree of degradation, it advances the processing to the next step. Figure 4 The step S170 shown.

[0243] Regarding the subsequent steps S170 to S190, the processing circuit 510 performs the same processing as in the first embodiment.

[0244] <Effect of the third embodiment>

[0245] The data center 500, as an information processing device in the third embodiment, extracts a portion of the raw data collected within a predetermined period using multiple sensors mounted on the vehicle 10. Using the extracted data, the data center 500 analyzes the extent of accumulated damage at the oil seal 51 located in the differential side 50.

[0246] Data center 500 includes a processing circuit 510 that performs processing according to a program. The raw data includes the rotational speed of a drive shaft 57, which is a rotating body and slides in contact with an oil seal 51 mounted on vehicle 10, as a first characteristic. The raw data includes the temperature of the fluid sealed by the oil seal 51, i.e., the ATF temperature, as a second characteristic. In this data center 500, the processing circuit 510 performs a search process. The search process includes a first process (step S130), which divides the data of the first characteristic into multiple parts according to the second characteristic included in the raw data, and calculates the frequency distribution of the raw data for each part regarding the first characteristic. The search process includes a second process (step S110), which sets multiple time windows to extract data from a portion of the raw data, such that the sum of the periods of all time windows is shorter than the overall period of the raw data. The search process includes a third process (step S120) to extract data from the raw data through the multiple time windows. The data obtained by combining all the data extracted through the multiple time windows is the extracted data. The search process includes a fourth process (step S130), which divides the data of the first feature into multiple partitions corresponding to the second feature of the original data, and calculates the frequency distribution of the extracted data with respect to the first feature for each partition. The search process also includes a fifth process (steps S140 and S150) which calculates the error between the frequency distribution of the original data and the frequency distribution of the extracted data to determine whether the original data and the extracted data are similar. After executing the first process, the processing circuit 510 performs a search process that repeatedly executes the second to fifth processes by changing the settings of multiple time windows. Then, the processing circuit 510 extracts the extracted data with an error below a threshold. Using the extracted data with an error below the threshold, the processing circuit 510 calculates the degradation degree as a damage index value (step S160).

[0247] According to the data center 500, the distribution of characteristic quantities related to the damage to the oil seal 51 can be analyzed using extracted data similar to the original data. Therefore, the data center 500 can obtain analysis results that are close to those obtained using the original data for damage analysis.

[0248] The extracted data obtained through Data Center 500 is obtained by cutting a portion of the original data. Therefore, the amount of extracted data is smaller compared to the original data. The more data used in parsing, the longer the processing time required for parsing the degree of corruption. By using extracted data, Data Center 500 is able to shorten the parsing time compared to using the original data.

[0249] <Effects of the Third Embodiment>

[0250] In addition to the effects of the first embodiment (1-1) to (1-5), the third embodiment also has the following effects.

[0251] (3-1) The data center 500 analyzes the degree of damage to the sealing component, i.e., the oil seal 51, which is in sliding contact with the drive shaft 57, which is a rotating body. The processing circuit 510 of the data center 500 sets the rotational speed of the drive shaft 57 as the first characteristic quantity and the ATF temperature as the second characteristic quantity.

[0252] The oil seal 51, which slides in contact with the drive shaft 57, wears and deteriorates due to friction with the drive shaft 57. Therefore, the higher the rotational speed of the drive shaft 57, the more easily the deterioration caused by friction is aggravated. Regarding the oil seal 51, the higher the temperature of the oil seal 51, the more easily the deterioration is aggravated. The temperature of the oil seal 51 is affected by the temperature of the fluid sealed by the oil seal 51, i.e., the ATF. The aforementioned data center 500 uses the above two physical quantities that affect the magnitude of damage accumulated at the oil seal 51 as feature quantities to obtain extracted data. Therefore, according to the aforementioned data center 500, data suitable for analyzing the degree of damage to the oil seal 51 can be extracted.

[0253] <Example of a modification to the third embodiment>

[0254] The third embodiment described above can be implemented with modifications as described below. The third embodiment and the following modifications to the third embodiment can be combined with each other without creating technical contradictions.

[0255] The aforementioned data center 500 divides the rotational speed of the drive shaft 57 into multiple values ​​based on the ATF temperature for data extraction. The data center 500 can use external air temperature as a second characteristic instead of ATF temperature. The processing circuit 510 of the data center 500 divides the rotational speed of the drive shaft 57 into multiple values ​​based on the external air temperature and calculates the frequency distribution of the original data and the extracted data. For example, Figure 22 The frequency distribution of the rotational speed of the drive shaft 57 in the raw data is shown when the external temperature around the vehicle 10, which is the object of analysis, is lower than a predetermined temperature. Figure 23 This diagram shows the frequency distribution of the rotational speed of the drive shaft 57 in the raw data when the external temperature around the vehicle 10, which is the object of analysis, is above a predetermined temperature. In these frequency distributions, zero rotational speed is set as the smallest level, and the rotational speed is divided into m levels from "1" to "m". For each division of the external temperature, the processing circuit 510 calculates... Figure 22 and Figure 23 The frequency distribution of the aforementioned rotational speeds in the original data and extracted data is shown. Data Center 500 is able to extract data from the original data using a frequency distribution calculated based on external temperature.

[0256] Oil seal 51 wears and deteriorates due to friction with drive shaft 57. Therefore, the higher the rotational speed of drive shaft 57, the more easily the deterioration caused by friction is aggravated. Regarding oil seal 51, the higher the temperature of oil seal 51, the more easily the deterioration is aggravated. The temperature of oil seal 51 is affected by the external air temperature. The aforementioned data center 500 uses the above two physical quantities that affect the aggravation of oil seal 51 as feature quantities to obtain extracted data. Therefore, based on the aforementioned data center 500, data suitable for analyzing the degree of damage to oil seal 51 can be extracted.

[0257] The aforementioned data center 500 extracts data using the rotational speed of the drive shaft 57 as the first characteristic quantity and the ATF temperature as the second characteristic quantity. However, the physical quantities used as the first and second characteristic quantities are not limited to the aforementioned physical quantities. Any physical quantity that affects the degree of damage to the sealing member that the rotating body slides in contact with can be used, such as humidity, climate information based on vehicle location information, etc. For example, humidity affects the deterioration of the oil seal 51. Regarding the oil seal 51, the higher the humidity around the seal, the more easily the deterioration is aggravated.

[0258] The data center 500 analyzes the degree of damage to the oil seal 51, which is a sealing member in sliding contact with a rotating body, located in the differential side 50. However, the sealing member whose degree of damage can be analyzed by the data center 500 is not limited to the oil seal 51. For example, any sealing member that is in sliding contact with a rotating body such as a transmission shaft or drive shaft can be used to analyze the degree of damage.

[0259] <Fourth Implementation>

[0260] Next, refer to Figures 24-34 This section describes a fourth embodiment of the information processing device. Furthermore, the fourth embodiment is an information processing device for analyzing the degree of damage to the planetary gear unit 61 mounted on a vehicle. The fourth embodiment differs from the first embodiment in that the device or component used to analyze the degree of damage is different. In the following description, the parts that differ from the first embodiment will be mainly explained. Detailed descriptions of components that are repeated in the first embodiment will be omitted. In the fourth embodiment, the information processing device for analyzing the degree of damage is also the processing circuit 510 of the data center 500.

[0261] <Structure of Power Distribution Mechanism 60>

[0262] The power distribution mechanism 60 is a power transmission device that transmits the power generated by the engine 21, the first electric generator 23A, and the second electric generator 23B to the drive wheels of the vehicle 10. The power distribution mechanism 60 is as follows... Figure 24 As shown, the housing 63 contains a planetary gear unit 61, a reduction gear 70, a motor gear 73, and a differential ring gear 74. Figure 24 The power distribution mechanism 60 shown is used, for example, in a front-wheel-drive vehicle. The planetary gear unit 61 includes a sun gear 66, three pinions 67, a planet carrier 64, and a ring 65. The sun gear 66 is located at the center of the planetary gear unit 61. The sun gear 66 is connected to a first electric generator 23A. The three pinions 67 are arranged around the sun gear 66, supported by the planet carrier 64. The ring 65 has a gear ring 68 on its inner circumferential surface and an engine output gear 69 on its outer circumferential surface. The rotation axes of the sun gear 66, the planet carrier 64, and the ring 65 are coaxial with the engine output shaft 22. The engine output shaft 22 is the output shaft of the engine 21. The output of the engine 21 is input to the planet carrier 64.

[0263] exist Figure 24 The diagram shows axes S1 to S4. Axis S1 is the axis through which the rotational shafts of the sun gear 66, planet carrier 64, and ring gear 65 pass, and the engine output shaft 22 passes. Axis S2 is the axis through which the rotational shaft of the reduction gear 70 passes. Axis S3 is the axis through which the output shaft of the second electric generator 23B and the rotational shaft of the motor gear 73 pass. The motor gear 73 is fixed to the output shaft of the second electric generator 23B. Axis S4 is the axis through which the rotational shaft of the differential ring gear 74 passes.

[0264] The output torque of engine 21 is input to planetary gear unit 61 via engine output shaft 22. This input torque is distributed from pinion 67, connected by planetary carrier 64, to sun gear 66 and ring gear 68. Sun gear 66 is connected to first electric generator 23A. First electric generator 23A is, for example, a rotary machine used for both power generation and propulsion. The torque distributed to ring gear 68 drives ring 65 to rotate, thereby driving engine output gear 69. Engine output gear 69 meshes with reduction gear 71 of reduction gear 70. Simultaneously, reduction gear 71 also meshes with motor gear 73 included in second electric generator 23B. Second electric generator 23B is, for example, a rotary machine for propulsion. Figure 24 As shown, the reduction gear 70 includes a large reduction gear 71 and a small reduction gear 72. The small reduction gear 72 meshes with the differential ring gear 74. The driving force of the differential ring gear 74 is transmitted to the drive wheels via the differential device 62. With the above structure, the power distribution mechanism 60 can transmit power to the drive wheels by combining the input torques of the engine 21 and the electric generator 23 into one through the large reduction gear 71.

[0265] Each gear component within the housing 63, including the planetary gear unit 61, is lubricated with lubricating oil. Lubrication by lubricating oil is performed, for example, by using gears immersed in lubricating oil to gather and disperse the lubricating oil supplied by an oil pump. The lubricating oil accumulates in the lower part of the housing 63. Figure 24 The single-dotted line indicates the level of lubricating oil accumulated within the casing 63. For example... Figure 24 As shown, the lower part of the gear of the differential ring gear 74 is immersed in lubricating oil. Therefore, when the differential ring gear 74 rotates, the gear inside the housing 63 is lubricated by the upward-swept lubricating oil. On the other hand, the first oil pump 75 and the second oil pump 76 draw in the lubricating oil accumulated in the lower part of the housing 63 and output it to the lubricating oil supply path. The first oil pump 75 is, for example, an oil pump connected to the gear driven by the differential ring gear 74. In this case, the lubricating oil supply implemented by the first oil pump 75 is related to the rotational speed of the differential ring gear 74. The second oil pump 76 is, for example, an oil pump connected to the engine output shaft 22 of the engine 21. In this case, the lubricating oil supply implemented by the second oil pump 76 is related to the engine speed of the engine 21. Therefore, the lubrication state of the planetary gear unit 61 is affected by the upward flow of lubricating oil implemented by the differential ring gear 74 and the supply of lubricating oil implemented by the first oil pump 75 and the second oil pump 76.

[0266] <Data Extraction>

[0267] The information processing terminal 600 sends an indication to the data center 500 regarding the extent of damage to the planetary gear unit 61. Then, similarly to the first embodiment, the data center 500 performs data extraction from the raw data through multiple time windows.

[0268] Figure 25 The raw data of characteristic quantities related to the planetary gear unit 61 of a particular vehicle 10 are shown. Figure 25 The raw data shown is a portion of the data equivalent to 100,000 hours from vehicle 10, which is the object of analysis. Figure 25 The raw data shown includes the torque input to the planetary gear unit 61, such as the torque input to the planet carrier 64, as characteristic quantities. This raw data also includes the temperature of the lubricating oil (ATF) that lubricates the planetary gear unit 61, the rotational speed of the pump that ejects the lubricating oil from the planetary gear unit 61 (i.e., the second oil pump 76), and the tilt angle of the vehicle 10. These characteristic quantities are physical quantities that are related to damage to the planetary gear unit 61. Figure 25 (a) shows the torque input to the planet carrier 64. Figure 25 (b) shows the ATF temperature. Figure 25 (c) shows the rotational speed of the second oil pump 76. Figure 25(d) shows the tilt angle of vehicle 10. Data center 500 identifies a cutting pattern for extracting features from the raw data containing the aforementioned feature quantities. Using the extracted data extracted based on the cutting pattern identified by data center 500, processing circuit 510 analyzes the accumulated damage at the planetary gear unit 61 of vehicle 10, which is the object of analysis.

[0269] <Search Processing for Slicing Patterns>

[0270] like Figure 4 As shown, the processing circuit 510 performs the same series of processes as in the first embodiment according to the program.

[0271] In step S100, the processing circuit 510 acquires raw data for a specific vehicle 10. The raw data includes data used to analyze the extent of damage to the planetary gear unit 61 of the vehicle 10, which is the subject of the analysis.

[0272] Next, in the processing of step S110, the processing circuit 510 targets... Figure 25 The raw data of the characteristic quantities related to the damage to the planetary gear unit 61 shown are, in the same manner as in the first embodiment, determined by setting multiple time windows to determine the cutting mode. The multiple time windows are set in such a way that the sum of the periods of all the time windows is shorter than the total period of the raw data.

[0273] In step S120, the processing circuit 510, similarly to the first embodiment, extracts data by cutting data through multiple time windows based on the cutting mode determined in step S110.

[0274] In step S130, the processing circuit 510 calculates the frequency distribution of characteristic quantities related to the damage to the planetary gear unit 61. In analyzing the degree of damage to the planetary gear unit 61, the first characteristic quantity is the torque input to the planet carrier 64. The second characteristic quantity is the ATF temperature. The processing circuit 510 divides the torque input to the planet carrier 64 into multiple values ​​according to the ATF temperature, and, similarly to the first embodiment, calculates the frequency distribution of the raw data and the extracted data obtained by combining all data extracted through multiple time windows. Figure 26 The frequency distribution of torque input to planetary carrier 64 is shown in the raw data when the ATF temperature of vehicle 10, the object of analysis, is below a predetermined temperature. Figure 27 The frequency distribution of the torque input to the planetary carrier 64 is shown in the raw data when the ATF temperature of the vehicle 10, the object of analysis, is above a predetermined temperature. For example... Figure 26 and Figure 27As shown, in these frequency distributions, zero input torque is set as the smallest level, and the input torque is divided into m levels from "1" to "m". The frequency distribution of the extracted data is also calculated based on the levels corresponding to the original data. In this example, the processing circuit 510 divides the ATF temperature into two categories: a category below a predetermined temperature and a category above a predetermined temperature, based on the torque input to the planetary carrier 64 contained in the original data and the extracted data. The processing circuit 510 calculates the frequency distribution described above for each of the two ATF temperature categories.

[0275] Next, the processing circuit 510 in Figure 4 In the processing of step S140 shown, similarly to the first embodiment, for each of the multiple partitions based on the second feature, the error between the frequency distribution of the first feature in the original data and the frequency distribution of the first feature in the extracted data is calculated. Regarding the error, for example, similar to the first embodiment, the formula for calculating the mean absolute error (MAE), i.e., Formula 1, can be used. In this case, if... Figure 26 and Figure 27 In the example shown, "n" is "m". Similarly, "i" is the number from "1" to "m". If the error is calculated with respect to the entire division using Formula 1 above, then the processing circuit 510 advances the processing to step S150.

[0276] The processing in step S150 is the same as that performed in the first embodiment. The processing circuit 510 uses the error of each segment to determine whether the original data and the extracted data are similar. Then, the processing circuit 510 repeatedly performs steps S110 to S150 by changing the settings of multiple time windows until it can extract data with an error below a threshold. As a result, the storage device 520 stores the cutting patterns where each error is below the threshold. In this way, the processing circuit 510 obtains a cutting pattern for extracted data similar to the original data.

[0277] <Calculation of fatigue damage and notification of failure prediction>

[0278] In step S160, the processing circuit 510 extracts data from the original data based on a data extraction pattern similar to the original data stored in the storage device 520 after processing prior to step S150.

[0279] In the fourth embodiment, similar to the first embodiment, fatigue damage is calculated based on the extracted data as an index value representing the degree of damage accumulated at the planetary gear unit 61.

[0280] The planetary gear unit 61 experiences wear and cumulative damage due to repeated engagement and collisions between the gears. The greater the torque input to the planet carrier 64, the greater the stress exerted between the gears, and the greater the cumulative damage due to engagement and collisions. The lubrication condition of the planetary gear unit 61 affects the friction between the gears. If the temperature of the lubricating oil in the planetary gear unit 61 increases, the lubricating oil supply to the planetary gear unit 61 becomes better, reducing friction between the gears. Therefore, increasing the temperature of the lubricating oil in the planetary gear unit 61 inhibits the accumulation of damage to the planetary gear unit 61.

[0281] As an example, the processing circuit 510 calculates the fatigue damage degree of the planetary gear unit 61 using the following method.

[0282] The processing circuit 510 calculates the frequency distribution for each division based on data obtained by dividing the torque input to the planetary carrier 64 according to the current ATF temperature. The division with the lowest ATF temperature among the multiple divisions for which the frequency distribution is calculated is determined as the reference division. Next, the processing circuit 510 performs a process to correct the frequency distribution of divisions other than the division with the lowest ATF temperature to a frequency distribution comparable to that of the division with the lowest ATF temperature. Figure 28 This is a graph showing the relationship between the temperature division of the ATF and the weighted average of the input torque frequencies. (This is achieved through...) Figure 28 This allows the frequency distribution of a certain partition to be corrected to a frequency distribution comparable to that of the partition with the lowest ATF temperature. Figure 28 In this embodiment, the lowest temperature is designated as 'a', and the ATF temperature is divided into five categories from a to e. For example, when the frequency distribution in temperature category c is corrected to be equivalent to the frequency distribution in temperature category a, the frequency of each level of the corrected frequency distribution is calculated using Formula 4 in the same manner as in the third embodiment.

[0283] In Formula 4 above, "i" is the number that determines the rank in the frequency distribution. If it is... Figure 26 and Figure 27 In the example shown, "i" is the number from "1" to "m". "fi" is the frequency of level i in temperature division c. "Fi" is the frequency of level i when fi is corrected to the frequency in temperature division a. "C3" and "C1" are as follows: Figure 28 The figures shown are the coefficients used in the weighting of temperature division c and temperature division a, respectively.

[0284] Processing circuit 510 uses Formula 4 above to calculate the frequency of each level of the corrected frequency distribution. Then, processing circuit 510 adds the calculated frequency of each level to the frequency of each level corresponding to the frequency distribution with the lowest ATF temperature. In this way, processing circuit 510 merges multiple frequency distributions into the frequency distribution with the lowest ATF temperature, and calculates the new frequency distribution, i.e., the corrected frequency distribution, from the extracted data.

[0285] Next, the processing circuit 510 calculates the fatigue damage degree based on the corrected frequency distribution. Figure 29 The diagram illustrates an example of a corrected frequency distribution in the analysis of the degree of damage to the planetary gear unit 61. In this corrected frequency distribution, the corrected input torque Tc is classified into six levels, A through F. The frequency Hij for each level represents the number of data points where the corrected input torque Tc is Ti or higher and lower than Tj. For example, data points where the corrected input torque Tc is T2 or higher and lower than T3 are classified into level B, and their frequency is represented as H23. An upper limit frequency Gij is determined for each level. The upper limit frequency Gij represents the maximum number of accumulated damages that would result in fatigue failure at the planetary gear unit 61 when damage caused by the corrected input torque Tc included in the corresponding level is accumulated. As an example, if L times of the torque included in the range where the corrected input torque Tc is T3 or higher and lower than T4 are input to the planet carrier 64, fatigue failure will occur at the planetary gear unit 61. Processing circuit 510 uses the frequency Hij for each level and the upper limit frequency Gij for each level in the corrected frequency distribution to calculate the fatigue damage degree according to Formula 3, similar to the first embodiment. If the processing circuit 510 calculates the fatigue damage degree, it advances the processing to the next step. Figure 4 The step S170 shown.

[0286] Regarding the subsequent steps S170 to S190, the processing circuit 510 performs the same processing as in the first embodiment.

[0287] <Function of the Fourth Embodiment>

[0288] The data center 500 of the information processing device in the fourth embodiment extracts a portion of the raw data collected over a predetermined period using multiple sensors mounted on the vehicle 10, and analyzes the extent of damage accumulated at the planetary gear unit 61.

[0289] Data center 500 includes a processing circuit 510 that performs processing according to a program. The raw data includes the torque input to the planetary gear unit 61 mounted on vehicle 10, i.e., the torque input to the planet carrier 64, as a first characteristic quantity. The raw data includes the temperature of the lubricating oil that lubricates the planetary gear unit 61, i.e., the ATF temperature, as a second characteristic quantity. In this data center 500, the processing circuit 510 performs a search process. The search process includes a first process (step S130), which divides the data of the first characteristic quantity into multiple parts according to the second characteristic quantity included in the raw data, and calculates the frequency distribution of the raw data for each part regarding the first characteristic quantity. The search process includes a second process (step S110), which sets multiple time windows to extract data from a portion of the raw data, such that the sum of the periods of all time windows is shorter than the overall period of the raw data. The search process includes a third process (step S120) to extract data from the raw data through multiple time windows. The data obtained by combining all the data extracted through the multiple time windows is the extracted data. The search process includes a fourth process (step S130), which divides the data of the first feature quantity into multiple partitions corresponding to the second feature quantity of the original data, and calculates the frequency distribution of the extracted data with respect to the first feature quantity for each partition. The search process includes a fifth process (steps S140 and S150) which calculates the error between the frequency distribution of the original data and the frequency distribution of the extracted data to determine whether the original data and the extracted data are similar. After executing the first process, the processing circuit 510 performs a search process that repeatedly executes the second to fifth processes by changing the settings of multiple time windows. Then, the processing circuit 510 extracts the extracted data with an error below a threshold. Using the extracted data with an error below the threshold, the processing circuit 510 calculates the fatigue damage degree as an indicator value of damage (step S160).

[0290] According to the data center 500, the distribution of characteristic quantities related to the damage to the planetary gear unit 61 can be analyzed using extracted data similar to the original data. Therefore, the data center 500 can obtain analysis results that are close to those obtained using the original data for damage analysis.

[0291] The extracted data obtained through Data Center 500 is obtained by cutting a portion of the original data. Therefore, the amount of extracted data is smaller compared to the original data. The more data used in parsing, the longer the processing time required for parsing the degree of corruption. By using extracted data, Data Center 500 is able to shorten the parsing time compared to using the original data.

[0292] <Effects of the Fourth Implementation>

[0293] In addition to the effects of the first embodiment (1-1) to (1-5), the fourth embodiment also has the following effects.

[0294] (4-1) Data center 500 analyzes the degree of damage to planetary gear unit 61. Data center 500's processing circuit 510 sets the torque input to planetary carrier 64 as the first characteristic quantity and the ATF temperature as the second characteristic quantity.

[0295] The planetary gear unit 61 experiences wear and cumulative damage due to repeated engagement and collisions between its gears. The greater the torque input to the planet carrier 64, the greater the stress exerted between the gears, and the greater the cumulative damage due to engagement and collisions. The lubrication condition of the planetary gear unit 61 affects the friction exerted between its gears. If the temperature of the lubricating oil in the planetary gear unit 61 increases, the supply of lubricating oil to the planetary gear unit 61 becomes better, reducing friction between the gears. Therefore, if the temperature of the lubricating oil in the planetary gear unit 61 increases, the accumulation of damage to the planetary gear unit 61 is suppressed. The data center 500 uses these two physical quantities affecting the accumulation of damage to the planetary gear unit 61 as feature quantities to obtain extracted data. Therefore, based on the data center 500, data suitable for analyzing the degree of damage to the planetary gear unit 61 can be extracted.

[0296] <Example of a modification to the fourth embodiment>

[0297] The fourth embodiment described above can be implemented by modifications as described below. The fourth embodiment and the following modifications to the fourth embodiment can be combined with each other to implement them without creating technical inconsistencies.

[0298] The aforementioned data center 500 divides the torque input to the planetary carrier 64 into multiple portions based on the ATF temperature for data extraction. The data center 500 can use the rotational speed of the pump dispensing lubricating oil to lubricate the planetary gear unit 61 as a second characteristic quantity, instead of the ATF temperature. The processing circuit 510 of the data center 500 divides the torque input to the planetary carrier 64 into multiple portions based on the rotational speed of the pump dispensing lubricating oil to lubricate the planetary gear unit 61, and calculates the frequency distribution of the raw data and the extracted data. The aforementioned pump is, for example, a second oil pump 76. For example, Figure 30 The frequency distribution of the torque input to the planetary carrier 64 is shown in the raw data when the rotational speed of the second oil pump 76 of the vehicle 10, which is the subject of analysis, is lower than a predetermined speed. Figure 31The diagram shows the frequency distribution of the torque input to the planetary carrier 64 in the raw data when the rotational speed of the second oil pump 76 of the vehicle 10, which is the object of analysis, is above a predetermined speed. In these frequency distributions, the input torque is divided into m levels, from "1" to "m", with zero as the minimum level. The processing circuit 510 calculates... for each division of the rotational speed of the second oil pump 76. Figure 30 and Figure 31 The frequency distribution of the aforementioned input torque in the raw data and extracted data is shown. The data center 500 is able to extract the extracted data from the raw data using the frequency distribution calculated according to the rotational speed of the second oil pump 76.

[0299] The lubrication condition of the planetary gear unit 61 affects the friction between the gears operating within it. If the rotational speed of the second oil pump 76 increases, the supply of lubricating oil to the planetary gear unit 61 becomes better, reducing friction between the gears. Therefore, the higher the rotational speed of the second oil pump 76, the more effectively the accumulation of damage to the planetary gear unit 61 is suppressed. The data center 500 uses these two physical quantities affecting the accumulation of damage to the planetary gear unit 61 as feature quantities to obtain extracted data. Therefore, based on the data center 500, data suitable for analyzing the degree of damage to the planetary gear unit 61 can be extracted.

[0300] On the other hand, the data center 500 can use the tilt angle of the vehicle 10 as a second characteristic quantity instead of the ATF temperature. The processing circuit 510 of the data center 500 divides the torque input to the planetary carrier 64 into multiple values ​​according to the tilt angle of the vehicle 10, and calculates the frequency distribution of the raw and extracted data. For example, Figure 32 The frequency distribution of the torque input to the planetary carrier 64 is shown in the raw data when the tilt angle of the vehicle 10, which is the object of analysis, is positive. When the tilt angle is positive, the vehicle 10 is on an uphill road. Figure 33 The frequency distribution of the torque input to the planetary carrier 64 is shown in the raw data when the tilt angle of the vehicle 10, which is the object of analysis, is zero. Figure 34 This diagram shows the frequency distribution of the torque input to the planetary carrier 64 in the raw data when the tilt angle of the vehicle 10, the object of analysis, is negative. When the tilt angle is negative, the vehicle 10 is on a downhill slope. In these frequency distributions, the input torque is divided into m levels, from "1" to "m", with zero as the minimum level. The processing circuit 510 calculates... for each division of the vehicle 10's tilt angle... Figures 32-34 The frequency distribution of the aforementioned input torque in the raw and extracted data is shown. The data center 500 is able to extract the extracted data from the raw data using the frequency distribution calculated according to the tilt angle of the vehicle 10.

[0301] The planetary gear unit 61 experiences wear and cumulative damage due to repeated engagement and collisions between its gears. The greater the torque input to the planetary carrier 64, the greater the stress exerted between the gears, and the greater the cumulative damage due to engagement and collisions. The lubrication state of the planetary gear unit 61 affects the friction exerted between its gears. When the vehicle 10 tilts, the level of the lubricating oil in the housing 63 changes, altering the immersion state of the differential ring gear 74 immersed in the lubricating oil. Therefore, due to the tilt of the vehicle 10, the lubrication state of the planetary gear unit 61, achieved by the upward movement of the lubricating oil, changes, affecting the accumulated damage at the planetary gear unit 61. The data center 500 uses these two physical quantities affecting the accumulation of damage to the planetary gear unit 61 as feature quantities to obtain extracted data. Therefore, based on the data center 500, data suitable for analyzing the degree of damage to the planetary gear unit 61 can be extracted.

[0302] The data center 500 extracts data by using the torque input to the planetary gear unit 61 as the first characteristic quantity and the ATF temperature as the second characteristic quantity. However, the physical quantities used as the first and second characteristic quantities are not limited to those mentioned above. Any physical quantity other than those mentioned above can be used as long as it affects the degree of damage to the planetary gear unit 61. For example, physical quantities reflecting the vehicle's driving mode, such as hybrid mode or EV mode, can also be used. In EV mode, the second oil pump 76, which rotates along with the engine 21, stops, and the supply of lubricating oil to the planetary gear unit 61 becomes poor. Therefore, during driving in EV mode, the cumulative damage to the planetary gear unit 61 increases.

[0303] The data center 500 described above analyzes the extent of damage to the planetary gear unit 61, which is formed by the meshing of the sun gear 66, pinion 67, and ring gear 68. However, the gear components whose extent of damage can be analyzed by the data center 500 are not limited to the planetary gear unit 61 described above. Similar to the planetary gear unit 61, it can be used to analyze the extent of damage to other gear components and power distribution devices formed by the meshing of multiple gears.

[0304] <Fifth Implementation>

[0305] Next, refer to Figures 35-40This section describes a fifth embodiment of the information processing device. Furthermore, the fifth embodiment is an information processing device for analyzing the degree of damage to the external connector 80B of the drive shaft 80 mounted on a vehicle. The fifth embodiment differs from the first embodiment in that the device or component that analyzes the degree of damage is different. In the following description, the parts that differ from the first embodiment will be mainly explained. Detailed descriptions of components that are repeated in the first embodiment are omitted. In the fifth embodiment, the information processing device for analyzing the degree of damage is also the processing circuit 510 of the data center 500.

[0306] <Structure of Steering Mechanism 81>

[0307] Figure 35 This is a schematic diagram showing the steering mechanism 81 of vehicle 10. Vehicle 10 is a front-wheel drive automobile where the front wheels serve as both drive wheels and steering wheels. The steering mechanism 81 uses power from the engine 21 and electric generator 23 transmitted via the power distribution mechanism 60 to drive the steering drive wheels 86. The steering mechanism 81 changes the steering angle of the steering drive wheels 86 according to the steering wheel angle. The drive shaft 80 transmits power from the power distribution mechanism 60 to the steering drive wheels 86. The differential device 62 of the power distribution mechanism 60 is connected to the left and right drive shafts 80. At the left and right drive shafts 80, an inner connector 80A is provided at the connection point with the differential device 62. In addition, an outer connector 80B is provided at the connection point with the steering drive wheels 86 on the left and right drive shafts 80. In vehicle 10, the inner connector 80A is a sliding constant velocity connector. The outer connector 80B is a fixed constant velocity connector. On the other hand, the steering angle of the steering drive wheel 86 is changed via the steering tie rod 82, steering arm 83, steering mechanism 84, and steering wheel 85. The left and right steering drive wheels 86 are held by the left and right steering arms 83, respectively. By connecting the left and right steering arms 83 with the steering tie rod 82, the left and right steering drive wheels 86 are integrated, causing the steering angle to change. The steering mechanism 84 is a device that converts the rotation of the steering wheel 85 into the movement of the steering tie rod 82. Thus, the rotational operation of the steering wheel 85 is reflected as a change in the steering angle of the left and right steering drive wheels 86 via the steering tie rod 82 and the left and right steering arms 83. At this time, at the outer joint 80B, the connection angle between the steering drive wheel 86 and the drive shaft 80, i.e., the joint angle, is generated according to the steering angle of the steering drive wheel 86.

[0308] The external connector 80B, located at the connection between the drive shaft 80 and the steering drive wheel 86, is an example of a site where fatigue damage occurs at the drive shaft 80. Fatigue damage to the external connector 80B can occur, for example, on the surface layer of the ball groove of the external connector 80B.

[0309] When vehicle 10 is in motion, the torque output from engine 21 and electric generator 23 is input to drive shaft 80 via power distribution mechanism 60. The torque input to drive shaft 80 is transmitted to steering drive wheel 86 via external connector 80B. Due to this input torque, stress is generated at external connector 80B, and fatigue damage occurs at external connector 80B due to the load generated by this stress. In addition, the magnitude of the load caused by the stress generated at external connector 80B by the aforementioned input torque is affected by the joint angle of external connector 80B.

[0310] <Data Extraction>

[0311] The information processing terminal 600 sends an indication to the data center 500 regarding the degree of damage to the analytical drive shaft 80. Then, similarly to the first embodiment, the data center 500 performs data extraction from the raw data through multiple time windows.

[0312] Figure 36 The raw data of characteristic quantities related to the drive shaft 80 of a particular vehicle 10 are shown. Figure 36 The raw data shown is a portion of the data equivalent to 100,000 hours from vehicle 10, which is the object of analysis. Figure 36 The raw data shown includes the torque input to the drive shaft 80 and the steering wheel angle as characteristic quantities. These characteristic quantities are physical quantities that are correlated with damage to the drive shaft 80. Figure 36 (a) shows the torque input to the drive shaft 80. Figure 36 (b) shows the steering wheel angle. Regarding the steering wheel angle, facing forward of the vehicle 10, the right-hand steering angle is set to positive, and the left-hand steering angle is set to negative. The data center 500 identifies a cutting pattern for extracting features from the raw data containing the aforementioned characteristic quantities. Using the extracted data extracted based on the cutting pattern identified by the data center 500, the processing circuit 510 analyzes the accumulated damage at the drive shaft 80 of the vehicle 10, which is the object of analysis.

[0313] <Search Processing for Slicing Patterns>

[0314] like Figure 4 As shown, the processing circuit 510 performs the same series of processes as in the first embodiment according to the program.

[0315] In step S100, the processing circuit 510 acquires raw data for a specific vehicle 10. The raw data includes data used to analyze the extent of damage to the drive shaft 80 of the vehicle 10, which is the subject of the analysis.

[0316] Next, in the processing of step S110, the processing circuit 510 targets... Figure 36The raw data of the characteristic quantities related to the damage to the drive shaft 80 shown are, in the same manner as in the first embodiment, determined by setting multiple time windows to determine the cutting mode. The multiple time windows are set in such a way that the sum of the periods of all the time windows is shorter than the total period of the raw data.

[0317] In step S120, the processing circuit 510, similarly to the first embodiment, extracts data by cutting data through multiple time windows based on the cutting mode determined in step S110.

[0318] In step S130, the processing circuit 510 calculates the frequency distribution of characteristic quantities related to the damage to the drive shaft 80. In analyzing the degree of damage to the drive shaft 80, the first characteristic quantity is the torque input to the drive shaft 80. The second characteristic quantity is the steering wheel angle. The processing circuit 510 divides the torque input to the drive shaft 80 into multiple components according to the steering wheel angle, and similarly to the first embodiment, calculates the frequency distribution of the original data and the extracted data obtained by combining all data extracted through multiple time windows. Figure 37 The frequency distribution of the torque input to the drive shaft 80 is shown in the raw data when the steering wheel angle of the vehicle 10, which is the object of analysis, is above a predetermined angle to the right. Figure 38 The frequency distribution of the torque input to the drive shaft 80 is shown in the raw data when the steering wheel angle of the vehicle 10, which is the object of analysis, is lower than a predetermined angle to the right and left. Figure 39 This shows the frequency distribution of the torque input to the drive shaft 80 in the raw data when the steering wheel angle of the vehicle 10, which is the object of analysis, is above a predetermined angle to the left. For example... Figures 37-39 As shown, in these frequency distributions, zero input torque is set as the minimum level, and the input torque is divided into m levels from "1" to "m". The frequency distribution of the extracted data is also calculated based on the levels corresponding to the original data. In this example, the processing circuit 510 divides the torque input to the drive shaft 80 contained in the original data and the extracted data into three categories. Each category is a steering wheel angle of more than a predetermined angle to the right, a steering wheel angle of less than a predetermined angle to the right and left, and a steering wheel angle of more than a predetermined angle to the left. The processing circuit 510 calculates the frequency distribution as described above for each of the three steering wheel angle categories.

[0319] Next, the processing circuit 510 in Figure 4In the processing of step S140 shown, similarly to the first embodiment, for each of the multiple partitions based on the second feature, the error between the frequency distribution of the first feature in the original data and the frequency distribution of the first feature in the extracted data is calculated. Regarding the error, for example, similar to the first embodiment, the formula for calculating the mean absolute error (MAE), i.e., Formula 1, can be used. In this case, if... Figures 37-39 In the example shown, "n" is "m". Similarly, "i" is the number from "1" to "m". If the error is calculated with respect to the entire division using Formula 1 above, then the processing circuit 510 advances the processing to step S150.

[0320] The processing in step S150 is the same as that performed in the first embodiment. The processing circuit 510 uses the error of each segment to determine whether the original data and the extracted data are similar. Then, the processing circuit 510 repeatedly performs steps S110 to S150 by changing the settings of multiple time windows until it can extract data with an error below a threshold. As a result, the storage device 520 stores the cutting patterns where each error is below the threshold. In this way, the processing circuit 510 obtains a cutting pattern for extracted data similar to the original data.

[0321] <Calculation of fatigue damage and notification of failure prediction>

[0322] In step S160, the processing circuit 510 extracts data from the original data based on a data extraction pattern similar to the original data stored in the storage device 520 after processing prior to step S150.

[0323] In the fifth embodiment, similar to the first embodiment, fatigue damage is calculated based on the extracted data, serving as an indicator of the degree of damage accumulated at the drive shaft 80. The fatigue damage is calculated separately for the left and right drive shafts, respectively.

[0324] Regarding the drive shaft 80, due to driving, the outer connector 80B experiences wear and cumulative damage. The greater the torque input to the drive shaft 80, the greater the stress exerted on the mating head, resulting in greater cumulative damage at the drive shaft 80. Similarly, the greater the steering wheel angle, the greater the stress exerted on the outer connector 80B, leading to greater cumulative damage at the drive shaft 80.

[0325] As an example, the processing circuit 510 calculates the fatigue damage of the drive shaft 80 using the following method.

[0326] The processing circuit 510 calculates the frequency distribution for each division based on data obtained by dividing the torque input to the drive shaft 80 according to the current steering wheel angle. For data in divisions other than those where the steering wheel angle is lower than a predetermined angle to the right and left, the data on the torque input to the drive shaft 80 included in that division is corrected according to the steering wheel angle division. Regarding this correction, the joint angle of the outer joint 80B can be taken into account, and any method that can reflect the change in stress load caused by the joint angle can be used. As an example, if the corrected input torque Tc is calculated by defining a formula that incorporates the joint angle of the outer joint 80B, the subsequent processing is as follows.

[0327] Based on the corrected input torque Tc obtained using the formula described above, the processing circuit 510 aggregates all the divided frequency distributions into frequency distributions divided into those where the steering wheel angle is lower than a predetermined angle to the right and left, and calculates the new frequency distribution, i.e., the corrected frequency distribution, from the extracted data. Next, the processing circuit 510 calculates the fatigue damage degree based on the corrected frequency distribution. Figure 40 The diagram illustrates an example of a corrected frequency distribution in the analysis of the degree of damage to the drive shaft 80. In this corrected frequency distribution, the corrected input torque Tc is classified into six levels, A through F. The frequency Hij for each level represents the number of data points where the corrected input torque Tc is Ti or higher and lower than Tj. For example, data points where the corrected input torque Tc is T2 or higher and lower than T3 are classified into level B, and their frequency is represented as H23. An upper limit frequency Gij is determined for each level. The upper limit frequency Gij represents the maximum number of accumulated damages that would cause fatigue failure at the outer joint 80B of the drive shaft 80 if damage caused by the corrected input torque Tc included in the corresponding level is accumulated. As an example, if G34 is L times, and L times of torque included in the range where the corrected input torque Tc is T3 or higher and lower than T4 is input to the drive shaft 80, fatigue failure will occur at the outer joint 80B. Processing circuit 510 uses the frequency Hij for each level and the upper limit frequency Gij for each level in the corrected frequency distribution to calculate the fatigue damage degree according to Formula 3, similar to the first embodiment. If the processing circuit 510 calculates the fatigue damage degree, it advances the processing to the next step. Figure 4 The step S170 shown.

[0328] Regarding the subsequent steps S170 to S190, the processing circuit 510 performs the same processing as in the first embodiment.

[0329] <Function of the 5th Embodiment>

[0330] The data center 500 of the information processing device in the fifth embodiment extracts a portion of the raw data collected within a predetermined period using multiple sensors mounted on the vehicle 10, and analyzes the extent of damage accumulated at the drive shaft 80.

[0331] Data center 500 includes processing circuit 510 that performs processing according to a program. The raw data includes the torque input to the drive shaft 80 mounted on vehicle 10 as a first characteristic quantity. The raw data includes the steering wheel angle of vehicle 10 as a second characteristic quantity. In this data center 500, processing circuit 510 performs a search process. The search process includes a first process (step S130), in which the data of the first characteristic quantity is divided into multiple parts according to the second characteristic quantity included in the raw data, and a frequency distribution in the raw data for each part is calculated. The search process includes a second process (step S110), in which multiple time windows are set to extract data from a portion of the raw data, such that the sum of the periods of all time windows is shorter than the overall period of the raw data. The search process includes a third process (step S120) to extract data from the raw data through multiple time windows. The data obtained by combining all the data extracted through the multiple time windows is the extracted data. The search process includes a fourth process (step S130), which divides the data of the first feature quantity into multiple partitions corresponding to the second feature quantity of the original data, and calculates the frequency distribution of the extracted data with respect to the first feature quantity for each partition. The search process includes a fifth process (steps S140 and S150) which calculates the error between the frequency distribution of the original data and the frequency distribution of the extracted data to determine whether the original data and the extracted data are similar. After performing the first process, the processing circuit 510 performs a search process that repeatedly performs the second to fifth processes by changing the settings of multiple time windows. Then, the processing circuit 510 extracts the extracted data with an error below a threshold. The processing circuit 510 uses the extracted data with an error below the threshold to calculate the fatigue damage degree as an indicator value of the damage (step S160).

[0332] According to the data center 500, the distribution of characteristic quantities related to the damage to the drive shaft 80 can be analyzed using extracted data similar to the original data. Therefore, the data center 500 can obtain analysis results that are close to those obtained using the original data for damage analysis.

[0333] The extracted data obtained through Data Center 500 is obtained by cutting a portion of the original data. Therefore, the amount of extracted data is smaller compared to the original data. The more data used in parsing, the longer the processing time required for parsing the degree of corruption. By using extracted data, Data Center 500 is able to shorten the parsing time compared to using the original data.

[0334] <Effects of the 5th Embodiment>

[0335] In addition to the effects of the first embodiment (1-1) to (1-5), the fifth embodiment also has the following effects.

[0336] (5-1) Data center 500 analyzes the degree of damage to drive shaft 80. Data center 500's processing circuit 510 sets the torque input to drive shaft 80 as the first characteristic quantity and the steering wheel angle as the second characteristic quantity.

[0337] Regarding the drive shaft 80, due to driving, the outer connector 80B experiences wear and cumulative damage. The greater the torque input to the drive shaft 80, the greater the stress exerted on the mating head, resulting in greater cumulative damage at the drive shaft 80. Similarly, the greater the steering wheel angle, the greater the stress exerted on the outer connector 80B, further contributing to the cumulative damage at the drive shaft 80. The aforementioned data center 500 uses these two physical quantities affecting the cumulative damage at the drive shaft 80 as feature quantities to obtain extracted data. Therefore, based on the aforementioned data center 500, data suitable for analyzing the degree of damage to the drive shaft 80 can be extracted.

[0338] <Example of a modification to the fifth embodiment>

[0339] The fifth embodiment described above can be implemented by modifications as described below. The fifth embodiment and the following modifications to the fifth embodiment can be combined with each other to implement them without creating technical inconsistencies.

[0340] The aforementioned data center 500 extracts data by using the torque input to the drive shaft 80 as the first characteristic quantity and the steering wheel angle of the vehicle 10 as the second characteristic quantity. However, the physical quantities used as the first and second characteristic quantities are not limited to the aforementioned physical quantities. Any physical quantity that affects the degree of damage to the drive shaft 80 can be used, such as the rotational speed of the drive shaft 80, the vertical acceleration of the vehicle 10, road surface information based on the position information of the vehicle 10, and climate information. For example, the higher the rotational speed of the drive shaft 80, the greater the damage accumulated at the outer joint 80B within a certain period of time. For example, the greater the vertical acceleration of the vehicle 10, the greater the damage accumulated at the outer joint 80B due to the stress load generated by the vertical bumps of the vehicle 10.

[0341] The device or component capable of applying the above-described analytical method is not limited to the drive shaft 80 of the steering drive wheel, which is mainly used for vehicle direction switching as described in Embodiment 5. It can also be used to analyze the degree of damage to the drive wheel that is assisted in vehicle direction switching via four-wheel steering, i.e., the auxiliary steering drive wheel. The auxiliary steering drive wheel mainly corresponds to the drive wheels of the vehicle 10 other than the front wheels. For example, the auxiliary steering drive wheel is the rear wheel of a rear-wheel drive vehicle or the rear wheel of a four-wheel drive vehicle. Regarding the auxiliary steering drive wheel, the steering angle of the wheel is slightly changed within a range smaller than that of the steering wheel or steering drive wheel, depending on the steering wheel angle. In one example, at vehicle speeds below a predetermined speed, the auxiliary steering drive wheel produces a steering angle opposite to that of the steering wheel or steering drive wheel. At vehicle speeds above the predetermined speed, the auxiliary steering drive wheel produces a steering angle in the same direction as the steering wheel or steering drive wheel. The data center 500 can analyze the accumulated damage at the drive shaft 80 of such auxiliary steering drive wheels.

[0342] <Sixth Implementation>

[0343] Next, refer to Figures 41-52 The sixth embodiment of the information processing device will be described below. Furthermore, the sixth embodiment is an information processing device for analyzing the degree of damage to the battery 25 mounted in a vehicle. The sixth embodiment differs from the first embodiment in that the device or component used for analyzing the degree of damage is different. In the following description, the parts that differ from the first embodiment will be mainly explained. Detailed descriptions of components that are repeated in the first embodiment will be omitted. In the sixth embodiment, the information processing device for analyzing the degree of damage is also the processing circuit 510 of the data center 500.

[0344] <Structure of Battery 25 and PCU24>

[0345] For reference Figure 1 As explained, vehicle 10 is equipped with battery 25 and PCU24. For example... Figure 41 As shown, battery 25 is a battery pack comprising multiple individual cells 26. Each individual cell 26 is the smallest structural unit of battery 25 that functions as a rechargeable battery. Figure 41The battery 25 shown has six individual cells 26. Battery 25 is, for example, a lithium-ion battery. Battery 25 is connected to PCU 24. PCU 24 includes a converter 27, a first inverter 28A, and a second inverter 28B. The first inverter 28A and the second inverter 28B convert the power supplied from battery 25 to generator 23. The first inverter 28A converts the direct current supplied from battery 25 into a suitable alternating current and supplies it to the first generator 23A. The second inverter 28B converts the direct current supplied from battery 25 into a suitable alternating current and supplies it to the second generator 23B. On the other hand, converter 27 converts the power supplied from generator 23 to battery 25. Converter 27 converts the alternating current generated by generator 23 into direct current capable of storing electricity in battery 25.

[0346] For example, when power is supplied to the electric generator 23, the battery 25 discharges to enable the generator 23, which functions as a motor, to rotate. When the electric generator 23 performs regenerative power generation, the battery 25 receives this power supply and is charged. Through this repeated charging and discharging, the battery 25 deteriorates and accumulates damage. The damage accumulated at the battery 25 is affected by the battery temperature and the SOC (state of charge), which indicates the state of charge of the battery 25. For example, when the battery temperature is above a certain level, the higher the battery temperature, the more severe the degradation. For example, when the battery 25 is below a certain SOC value, degradation is more likely to accelerate. For example, when the battery 25 is above a certain SOC value, degradation is more likely to accelerate. To suppress the acceleration of such degradation, Figure 1 The second control device 92 shown can control the PCU 24 to limit the charging and discharging of the battery 25. If the battery 25 is charged and discharged using high power, its degradation will accelerate. To suppress such degradation, the second control device 92 can limit the power used for charging and discharging the battery 25. The upper limit value for the current used to charge the battery 25 is called the charging power upper limit value Win. The upper limit value for the current used to discharge the battery 25 is called the discharging power upper limit value Wout. For example, the second control device 92 can set the charging power upper limit value Win and the discharging power upper limit value Wout based on the battery temperature and SOC. The second control device 92 can also set the charging power upper limit value Win and the discharging power upper limit value Wout based on the degradation of the battery 25. The second control device 92 can also be set such that the greater the accumulated damage at the battery 25, the smaller the values ​​of the charging power upper limit value Win and the discharging power upper limit value Wout. In this case, the charging power upper limit value Win and the discharging power upper limit value Wout reflect the accumulated damage at the battery 25.

[0347] <Data Extraction>

[0348] The information processing terminal 600 sends an indication to the data center 500 regarding the extent of damage to the battery 25, which exchanges power with the first electric generator 23A, which functions as a motor. Then, similarly to the first embodiment, the data center 500 performs data processing by extracting data from the raw data through multiple time windows.

[0349] Figure 42 Raw data showing the characteristic quantities related to the battery 25 of a particular vehicle 10. Figure 42 The raw data shown is a portion of the data equivalent to 100,000 hours from vehicle 10, which is the object of analysis. Figure 42 The raw data shown includes the output of the first electric generator 23A, the temperature of the battery 25, the state of charge (SOC) of the battery 25, the upper limit of the charging power Win and the upper limit of the discharging power Wout of the battery 25 as characteristic quantities. These characteristic quantities are physical quantities that are related to damage to the battery 25. Figure 42 (a) shows the output of the first electric generator 23A. Figure 42 (b) shows the temperature of battery 25. Figure 42 (c) shows the SOC of battery 25. Figure 42 (d) indicates the upper limit of the charging power of battery 25, Win. Figure 42 (e) shows the upper limit of the discharge power of battery 25, Wout. Data center 500 identifies a cutting pattern for extracting features from the raw data containing the aforementioned feature quantities. Using the extracted data extracted based on the cutting pattern identified by data center 500, processing circuit 510 analyzes the damage accumulated at the battery 25 of vehicle 10, which is the object of analysis.

[0350] <Search Processing for Slicing Patterns>

[0351] like Figure 4 As shown, the processing circuit 510 performs the same series of processes as in the first embodiment according to the program.

[0352] In step S100, the processing circuit 510 acquires raw data for a specific vehicle 10. The raw data includes data used to analyze the extent of damage to the battery 25 of the vehicle 10 being analyzed.

[0353] Next, in the processing of step S110, the processing circuit 510 targets... Figure 42 The raw data of the characteristic quantities related to the damage to battery 25 shown are, in the same manner as in the first embodiment, determined by setting multiple time windows to determine the cutting mode. The multiple time windows are set in such a way that the sum of the periods of all the time windows is shorter than the total period of the raw data.

[0354] In step S120, the processing circuit 510, similarly to the first embodiment, extracts data by cutting data through multiple time windows based on the cutting mode determined in step S110.

[0355] In step S130, the processing circuit 510 calculates the frequency distribution of characteristic quantities related to the damage to the battery 25. In analyzing the degree of damage to the battery 25, the first characteristic quantity is the output of the first electric generator 23A. The second characteristic quantity is the temperature of the battery 25. The temperature of the battery 25 can, for example, be set as the average temperature of the six individual cells 26 constituting the battery 25. The processing circuit 510 divides the output of the first electric generator 23A into multiple portions according to the temperature of the battery 25, and, similarly to the first embodiment, calculates the frequency distribution of the raw data and the extracted data obtained by combining all data extracted through multiple time windows. Figure 43 The frequency distribution of the output of the first electric generator 23A is shown in the raw data when the temperature of the battery 25 of the vehicle 10, which is the subject of analysis, is lower than a predetermined temperature. Figure 44 The frequency distribution of the output of the first electric generator 23A in the raw data is shown when the temperature of the battery 25 of the vehicle 10, which is the subject of analysis, is above a predetermined temperature. For example... Figure 43 and Figure 44 As shown, in these frequency distributions, the levels are centered at zero, with an equal number of positive and negative levels. The positive output range is when the first electric generator 23A functions as a driving motor and discharges from the battery 25. The negative output range is when the first electric generator 23A functions as a generator and charges the battery 25. Figure 43 and Figure 44 In the example shown, the range of positive output and the range of negative output are each divided into "m" levels. In this example, the level with the smallest output value is set to "1", and the output is divided into 2m levels from "1" to "2m". The frequency distribution of the extracted data is also calculated by dividing it into levels corresponding to the original data. In this example, the processing circuit 510 divides the temperature of the battery 25 into two categories: a category below a predetermined temperature and a category above a predetermined temperature, based on the output of the first electric generator 23A included in the original data and the extracted data. The processing circuit 510 calculates the frequency distribution as described above for each of the two battery temperature categories.

[0356] Next, the processing circuit 510 in Figure 4In the processing of step S140 shown, similarly to the first embodiment, for each of the multiple partitions based on the second feature, the error between the frequency distribution of the first feature in the original data and the frequency distribution of the first feature in the extracted data is calculated. Regarding the error, for example, similar to the first embodiment, the formula for calculating the mean absolute error (MAE), i.e., Formula 1, can be used. In this case, if... Figure 43 and Figure 44 In the example shown, "n" is "2m". Similarly, "i" is the number from "1" to "2m". If the error is calculated with respect to the entire division using Formula 1 above, the processing circuit 510 advances the processing to step S150.

[0357] The processing in step S150 is the same as that performed in the first embodiment. The processing circuit 510 uses the error of each segment to determine whether the original data and the extracted data are similar. Then, the processing circuit 510 repeatedly performs steps S110 to S150 by changing the settings of multiple time windows until it can extract data with an error below a threshold. As a result, the storage device 520 stores the cutting patterns where each error is below the threshold. In this way, the processing circuit 510 obtains a cutting pattern for extracted data similar to the original data.

[0358] <Calculation of Degradation and Prediction of Failure>

[0359] In step S160, the processing circuit 510 extracts data from the original data based on a data extraction pattern similar to the original data stored in the storage device 520 after processing prior to step S150.

[0360] In the sixth embodiment, a degradation degree is calculated based on the extracted data, serving as an indicator of the extent of damage accumulated at battery 25. Regarding the degradation degree, for example, the degree of degradation leading to undesirable conditions related to battery 25 can be set to "1," and can be defined as a value from "0" to "1" representing the proportion of degradation accumulated at battery 25. Battery degradation refers, for example, to the degradation of the active materials inside the battery and an increase in the battery's internal resistance. Due to such degradation, undesirable conditions such as a decrease in charging and discharging capacity occur in the battery.

[0361] Battery 25 deteriorates and accumulates damage due to repeated charging and discharging. The greater the output of the first electric generator 23A that receives power from battery 25, the greater the discharge from battery 25. The greater the output of the first electric generator 23A that supplies power to battery 25, the greater the charging to battery 25. The greater the charging and discharging of battery 25, the more easily its deterioration is accelerated. Regarding battery 25, the higher its temperature, the more easily its deterioration is accelerated.

[0362] As an example, the processing circuit 510 calculates the degree of degradation of the battery 25 using the following method.

[0363] The processing circuit 510 calculates the frequency distribution for each division based on data obtained by dividing the output of the first electric generator 23A according to the temperature of the battery 25 at that time. One of the calculated divisions is determined as the reference division. Next, for the data in divisions other than the reference division, the output data included in that division is corrected according to the temperature division. Regarding this correction, the temperature of the battery 25 can be taken into account, and any method that can reflect the rate of change in the degradation response of the battery 25 caused by temperature can be used. As an example, if the corrected output Pc is calculated by defining a formula that incorporates the temperature of the battery 25, the subsequent processing is as follows.

[0364] Based on the corrected output Pc obtained using the formula described above, the processing circuit 510 aggregates all the divided frequency distributions into a single frequency distribution determined as a baseline, and calculates the new frequency distribution in the extracted data, i.e., the corrected frequency distribution. Next, the processing circuit 510 calculates the degradation degree based on the corrected frequency distribution. Figure 45 The diagram shows an example of a corrected frequency distribution in the analysis of the degree of damage to battery 25. In this corrected frequency distribution, the corrected output Pc is classified into eight levels, A through H. Levels A through D are four levels classifying the corrected output Pc when the first electric generator 23A functions as a generator. Levels E through H are four levels classifying the corrected output Pc when the first electric generator 23A functions as a driving motor. The frequency Hij for each level represents the number of data where the corrected output Pc value is Pi or higher and lower than Pj. For example, data where the corrected output Pc value is P2 or higher and lower than P3 are classified into level B, and their frequency is represented as H23. An upper limit frequency Gij is determined for each level. The upper limit frequency Gij represents the maximum number of accumulated damages that result in undesirable conditions due to degradation at battery 25 when damage caused by the corrected output Pc included in the corresponding level is accumulated. As an example, if an output of type L occurs where the corrected output Pc is above P3 and below P4, a deterioration-related condition occurs at battery 25. The processing circuit 510 calculates the degree of degradation using the frequency Hij for each level of the corrected frequency distribution and the upper limit frequency Gij for each level, similar to the third embodiment, according to the following formula 6.

[0365] [Formula 6]

[0366] Deterioration degree = H 12 / G 12 +H 23 / G 23 +…+H 78 / G 78 +H 89 / G 89

[0367] If the processing circuit 510 calculates the degree of degradation, it advances the processing to the next step. Figure 4 The step S170 shown.

[0368] Regarding the subsequent steps S170 to S190, the processing circuit 510 performs the same processing as in the first embodiment.

[0369] <Effect of the 6th Embodiment>

[0370] The data center 500 of the information processing device in the sixth embodiment extracts a portion of the raw data collected over a predetermined period using multiple sensors mounted on the vehicle 10, and analyzes the extent of damage accumulated at the battery 25.

[0371] Data center 500 includes processing circuit 510 that performs processing according to a program. The raw data includes the output of a first electric generator 23A, which is a motor mounted on vehicle 10, as a first characteristic. The raw data includes the temperature of battery 25 as a second characteristic. In this data center 500, processing circuit 510 performs a search process. The search process includes a first process (step S130), in which the data of the first characteristic is divided into multiple parts according to the second characteristic included in the raw data, and a frequency distribution in the raw data for each part is calculated. The search process includes a second process (step S110), in which multiple time windows are set to extract data from a portion of the raw data, such that the sum of the periods of all time windows is shorter than the overall period of the raw data. The search process includes a third process (step S120) to extract data from the raw data through the multiple time windows. The data obtained by combining all the data extracted through the multiple time windows is the extracted data. The search process includes a fourth process (step S130), which divides the data of the first feature into multiple partitions corresponding to the second feature of the original data, and calculates the frequency distribution of the extracted data with respect to the first feature for each partition. The search process also includes a fifth process (steps S140 and S150) which calculates the error between the frequency distribution of the original data and the frequency distribution of the extracted data to determine whether the original data and the extracted data are similar. After executing the first process, the processing circuit 510 performs a search process that repeatedly executes the second to fifth processes by changing the settings of multiple time windows. Then, the processing circuit 510 extracts the extracted data with an error below a threshold. Using the extracted data with an error below the threshold, the processing circuit 510 calculates the degradation degree as a damage index value (step S160).

[0372] According to the data center 500, the distribution of characteristic quantities related to the damage to battery 25 can be analyzed using extracted data similar to the original data. Therefore, the data center 500 can obtain analysis results that are close to those obtained using the original data for damage analysis.

[0373] The extracted data obtained through Data Center 500 is obtained by cutting a portion of the original data. Therefore, the amount of extracted data is smaller compared to the original data. The more data used in parsing, the longer the processing time required for parsing the degree of corruption. By using extracted data, Data Center 500 is able to shorten the parsing time compared to using the original data.

[0374] <Effects of the 6th Embodiment>

[0375] In addition to the effects of the first embodiment (1-1) to (1-5), the sixth embodiment also has the following effects.

[0376] (6-1) The data center 500 analyzes the extent of damage to the battery 25 that exchanges power with the first electric generator 23A, which acts as a motor. The processing circuit 510 of the data center 500 sets the output of the first electric generator 23A as a first characteristic quantity and the temperature of the battery 25 as a second characteristic quantity.

[0377] Battery 25 deteriorates and accumulates damage due to repeated charging and discharging. The greater the output of the first electric generator 23A, which receives power from battery 25, the greater the discharge from battery 25. The greater the output of the first electric generator 23A supplying power to battery 25, the greater the charging to battery 25. The greater the charging to and discharging from battery 25, the more easily battery 25 deteriorates. Regarding battery 25, the higher its temperature, the more easily deterioration is accelerated. The data center 500 uses these two physical quantities, which affect the magnitude of damage accumulated at battery 25, as feature quantities to obtain extracted data. Therefore, based on the data center 500, data suitable for analyzing the degree of damage to battery 25 can be extracted.

[0378] <Example of a modification to the sixth embodiment>

[0379] The sixth embodiment described above can be implemented by modifications as described below. The sixth embodiment and the following modifications to the sixth embodiment can be combined with each other to implement them without creating technical inconsistencies.

[0380] The data center 500 uses the average temperature of the six individual cells 26 constituting the battery 25 as the temperature of the battery 25. The temperature used by the data center 500 for the battery 25 is not limited to the average temperature of the six individual cells 26. The temperature of the individual cell 26 showing the highest temperature among the six individual cells 26 can be used as the temperature of the battery 25. The temperature of the individual cell 26 showing the lowest temperature among the six individual cells 26 can also be used as the temperature of the battery 25. In addition to the above, the temperatures of each individual cell 26 can be weighted considering the influence of each individual cell 26 on the overall temperature of the battery 25. For example, the average temperature can be calculated by weighting in a way that makes the temperature of the individual cells 26 located near the center of the battery 25 easily reflect the average temperature of the six individual cells 26.

[0381] The aforementioned data center 500 divides the output of the first electric generator 23A into multiple portions based on the temperature of the battery 25 for data extraction. The data center 500 can use the state of charge (SOC) of the battery 25 as a second characteristic quantity instead of the battery temperature. The processing circuit 510 of the data center 500 divides the output of the first electric generator 23A into multiple portions based on the SOC of the battery 25 and calculates the frequency distribution of the original data and the extracted data. For example, Figure 46 The frequency distribution of the output of the first electric generator 23A in the raw data is shown when the SOC of the battery 25 of the vehicle 10, which is the subject of analysis, is lower than the first predetermined value. Figure 47 The frequency distribution of the output of the first electric generator 23A is shown in the raw data when the SOC of the battery 25 of the vehicle 10 being analyzed is above a first predetermined value and below a second predetermined value. Figure 48 This diagram shows the frequency distribution of the output of the first electric generator 23A in the raw data when the SOC of the battery 25 of the vehicle 10 being analyzed is above a second predetermined value. Furthermore, the first predetermined value is set to be less than the second predetermined value. In these frequency distributions, levels are established with zero output as the center, and the number of positive and negative levels are equal. The positive output range represents the state where the first electric generator 23A functions as a driving motor and discharges from the battery 25. The negative output range represents the state where the first electric generator 23A functions as a generator and charges the battery 25. Figures 46-48 In the example shown, the range of positive output and the range of negative output are each divided into "m" levels. In this example, the level with the smallest output value is set to "1", and the output is divided into 2m levels from "1" to "2m". The processing circuit 510 calculates for each division of the SOC of the battery 25. Figures 46-48 The frequency distribution of the above outputs in the original data and extracted data is shown. Data Center 500 can extract data from the original data using the frequency distribution calculated according to the SOC of battery 25.

[0382] Battery 25 deteriorates and accumulates damage due to repeated charging and discharging. The greater the output of the first electric generator 23A, which receives power from battery 25, the greater the discharge from battery 25. The greater the output of the first electric generator 23A supplying power to battery 25, the greater the charging to battery 25. The greater the charging to and discharging from battery 25, the more easily battery 25 deteriorates. The deterioration of battery 25 is affected by the state of charge (SOC) of battery 25. The data center 500 uses these two physical quantities, which affect the magnitude of accumulated damage at battery 25, as feature quantities to obtain extracted data. Therefore, based on the data center 500, data suitable for analyzing the degree of damage to battery 25 can be extracted.

[0383] The data center 500 can use the upper limit of the charging power Win of the battery 25 as the second characteristic quantity, instead of the battery temperature. The processing circuit 510 of the data center 500 divides the output of the first electric generator 23A into multiple values ​​according to the upper limit of the charging power Win of the battery 25, and calculates the frequency distribution of the raw and extracted data. For example, Figure 49 The frequency distribution of the output of the first electric generator 23A in the raw data is shown when the upper limit of the charging power Win of the battery 25 of the vehicle 10, which is the object of analysis, is lower than a predetermined value. Figure 50 This diagram shows the frequency distribution of the output of the first electric generator 23A in the raw data when the upper limit of the charging power Win of the battery 25 of the vehicle 10 being analyzed is above a predetermined value. In these frequency distributions, levels are centered at output zero, with an equal number of positive and negative levels. The positive output range represents the state where the first electric generator 23A functions as a driving motor and discharges from the battery 25. The negative output range represents the state where the first electric generator 23A functions as a generator and charges the battery 25. Figure 49 and Figure 50 In the example shown, the range of positive output and the range of negative output are each divided into "m" levels. In this example, the lowest output value is set to "1", and the output is divided into 2m levels from "1" to "2m". The processing circuit 510 calculates for each division of the upper limit value Win of the charging power of the battery 25. Figure 49 and Figure 50 The frequency distribution of the above outputs in the original data and extracted data are shown. Data center 500 can extract extracted data from the original data using the frequency distribution calculated based on the upper limit of the charging power Win of battery 25.

[0384] Battery 25 deteriorates and accumulates damage due to repeated charging and discharging. The greater the output of the first electric generator 23A, which receives power from battery 25, the greater the discharge from battery 25. The greater the output of the first electric generator 23A supplying power to battery 25, the greater the charging to battery 25. The greater the charging to and discharging from battery 25, the more easily battery 25 deteriorates. The upper limit value of charging power Win is a value set based on the temperature and SOC of battery 25. The upper limit value of charging power Win is changed according to the damage accumulated at battery 25 to suppress the deterioration of battery 25. Therefore, the deterioration of battery 25 is affected by the magnitude of the upper limit value of charging power Win. The data center 500 uses the above two physical quantities that affect the magnitude of the damage accumulated at battery 25 as feature quantities to obtain extracted data. Therefore, according to the data center 500, data suitable for analyzing the degree of damage to battery 25 can be extracted.

[0385] The data center 500 can use the upper limit of the discharge power of battery 25, Wout, as the second characteristic quantity, instead of the temperature of battery 25. The processing circuit 510 of the data center 500 divides the output of the first electric generator 23A into multiple values ​​according to the upper limit of the discharge power of battery 25, Wout, and calculates the frequency distribution of the raw and extracted data. For example, Figure 51 The frequency distribution of the output of the first electric generator 23A in the raw data is shown when the upper limit of the discharge power Wout of the battery 25 of the vehicle 10, which is the subject of analysis, is lower than a predetermined value. Figure 52 This diagram shows the frequency distribution of the output of the first electric generator 23A in the raw data when the upper limit of the discharge power Wout of the battery 25 of the vehicle 10 being analyzed is above a predetermined value. In these frequency distributions, levels are set with zero output as the center, and the number of positive and negative levels are equal. The positive output range represents the state where the first electric generator 23A functions as a driving motor and discharges from the battery 25. The negative output range represents the state where the first electric generator 23A functions as a generator and charges the battery 25. Figure 51 and Figure 52 In the example shown, the range of positive output and the range of negative output are each divided into "m" levels. In this example, the lowest output value is set to "1", and the output is divided into 2m levels from "1" to "2m". The processing circuit 510 calculates for each division of the upper limit of the discharge power value Wout of the battery 25. Figure 51 and Figure 52 The frequency distribution of the above outputs in the original data and extracted data is shown. Data center 500 can extract extracted data from the original data using the frequency distribution calculated based on the upper limit of the discharge power value Wout of battery 25.

[0386] Battery 25 deteriorates and accumulates damage due to repeated charging and discharging. The greater the output of the first electric generator 23A, which receives power from battery 25, the greater the discharge from battery 25. The greater the output of the first electric generator 23A supplying power to battery 25, the greater the charging to battery 25. The greater the charging to and discharging from battery 25, the more easily battery 25 deteriorates. The discharge power limit value Wout is a value set based on the temperature and SOC of battery 25. The discharge power limit value Wout is changed according to the damage accumulated at battery 25 to suppress battery 25 deterioration. Therefore, the deterioration of battery 25 is affected by the magnitude of the discharge power limit value Wout. The data center 500 uses the above two physical quantities that affect the magnitude of damage accumulated at battery 25 as feature quantities to obtain extracted data. Therefore, according to the data center 500, data suitable for analyzing the degree of damage to battery 25 can be extracted.

[0387] The data center 500 sets the output of the first electric generator 23A as the first characteristic quantity. Alternatively, the data center 500 may set the output of the second electric generator 23B as the first characteristic quantity instead of the output of the first electric generator 23A. Furthermore, the data center 500 can set the overall output of the electric generator 23 obtained by adding the outputs of the first electric generator 23A and the second electric generator 23B as the first characteristic quantity.

[0388] Battery 25 exchanges power with both the first electric generator 23A and the second electric generator 23B for charging and discharging. That is, the power charged to battery 25 and the power discharged from battery 25 are the superposition of the charging and discharging with respect to the first electric generator 23A and the charging and discharging with respect to the second electric generator 23B. Therefore, if the overall output of the electric generator 23 is set as the first characteristic quantity, the magnitude of the charging and discharging output to battery 25 can be more appropriately reflected.

[0389] <Seventh Implementation>

[0390] Next, refer to Figures 53-64This section describes a seventh embodiment of the information processing device. Furthermore, the seventh embodiment is an information processing device for analyzing the degree of damage to the radiator 313 of the cooling system 310 of the engine 21 mounted on the vehicle 10. The seventh embodiment differs from the first embodiment in that the device or component being analyzed for the degree of damage is different. In the following description, the parts that differ from the first embodiment will be mainly explained. Detailed descriptions of components that are repeated in the first embodiment are omitted. In the seventh embodiment, the information processing device for analyzing the degree of damage is also the processing circuit 510 of the data center 500.

[0391] <Structure of the cooling system 310 in engine 21>

[0392] like Figure 53 As shown, a cooling system 310 is provided at the engine 21 mounted on the vehicle 10. The cooling system 310 is a water-cooled cooling device that supplies cooling water to the water-cooled jacket of the engine 21 to cool the structural components of the engine 21.

[0393] The cooling system 310 includes a radiator 313, a water pump 314, and a water thermometer 315. The radiator 313 is a heat exchanger that cools the refrigerant, i.e., the cooling water, circulating in the cooling system 310. The cooling water is, for example, LLC (Long Life Coolant). The radiator 313 is, for example, an air-cooled heat exchanger. The water pump 314 is a pump that circulates the cooling water to the radiator 313.

[0394] The cooling system 310 includes a first cooling water passage 311 and a second cooling water passage 312 connecting the radiator 313 to the water-cooled jacket of the engine 21. A water pump 314 is installed in the first cooling water passage 311. The water pump 314 draws in cooling water from the radiator 313 side and sprays it to the water-cooled jacket side. The first cooling water passage 311 supplies cooling water cooled by the radiator 313 to the water-cooled jacket. The second cooling water passage 312 is a passage for returning cooling water that has passed through the water-cooled jacket to the radiator 313 for cooling purposes. Figure 53 As indicated by the middle arrow, the water circulates in the cooling system 310. A water thermometer 315 is installed in the second cooling water passage 312. The water thermometer 315 measures the temperature of the cooling water flowing into the radiator 313 through the second cooling water passage 312.

[0395] The radiator 313 functions as a dynamic damper. The radiator 313 is mounted on the vehicle 10 via a support 321 and an elastic body 322. The support 321 is part of the vehicle body of the vehicle 10. For example, the support 321 is part of the frame structure of the vehicle body located at the front of the vehicle 10. The elastic body 322 is an elastic member capable of stretching and contracting in the vertical direction of the vehicle 10. The elastic body 322 is, for example, a spring. The elastic body 322 is, for example, a rubber elastic member. Through the stretching and contraction of the elastic body 322, the radiator 313 can vibrate vertically, partially independently of the vertical vibration of the vehicle 10. The radiator 313 vibrates vertically due to the vibration of the engine 21, thereby weakening the vibration of the vehicle body caused by the vibration of the engine 21, and thus functions as a dynamic damper to suppress the vibration of the vehicle 10.

[0396] <Data Extraction>

[0397] The information processing terminal 600 sends an indication to the data center 500 regarding the extent of damage to the heat sink 313. Then, similarly to the first embodiment, the data center 500 performs data extraction from the raw data through multiple time windows.

[0398] Figure 54 The raw data show the characteristic quantities related to the radiator 313 of a particular vehicle 10. Figure 54 The raw data shown is a portion of the data equivalent to 100,000 hours from vehicle 10, which is the object of analysis. Figure 54 The raw data shown includes the acceleration of the vehicle 10 equipped with radiator 313, such as the sprung acceleration of vehicle 10, as a characteristic quantity. Sprung acceleration is the acceleration generated in the body portion of vehicle 10 that is higher than the suspension. Sprung acceleration is, for example, along... Figure 53 The above data includes the vertical acceleration of the vehicle 10. The raw data includes the temperature of the refrigerant, i.e., the coolant, flowing into the radiator 313. The raw data includes the rotational speed of the internal combustion engine 21 connected to the radiator 313, i.e., the crankshaft speed. The crankshaft speed is the number of revolutions per minute of the engine output shaft 22. The raw data includes the rotational speed of the pump 314 that circulates coolant to the radiator 313. The above characteristic quantities are physical quantities that are related to damage to the radiator 313.

[0399] Figure 54 (a) shows the sprung acceleration of vehicle 10. Figure 54 (b) shows the temperature of the cooling water. Figure 54 (c) indicates the crankshaft speed. Figure 54(d) indicates the rotational speed of the water pump 314. The data center 500 identifies a cutting pattern for extracting features from the raw data containing the aforementioned characteristic quantities. Using the extracted data extracted based on the cutting pattern identified by the data center 500, the processing circuit 510 analyzes the accumulated damage at the radiator 313 of the vehicle 10, which is the object of analysis.

[0400] <Search Processing for Slicing Patterns>

[0401] like Figure 4 As shown, the processing circuit 510 performs the same series of processes as in the first embodiment according to the program.

[0402] In step S100, the processing circuit 510 acquires raw data for a specific vehicle 10. The raw data includes data used to analyze the extent of damage to the radiator 313 of the vehicle 10 being analyzed.

[0403] Next, in the processing of step S110, the processing circuit 510 targets... Figure 54 The raw data of the characteristic quantities related to the damage to the heat sink 313 shown are, in the same manner as in the first embodiment, determined by setting multiple time windows to determine the cutting mode. The multiple time windows are set in such a way that the sum of the periods of all the time windows is shorter than the total period of the raw data.

[0404] In step S120, the processing circuit 510, similarly to the first embodiment, extracts data by cutting data through multiple time windows based on the cutting mode determined in step S110.

[0405] In step S130, the processing circuit 510 calculates the frequency distribution of characteristic quantities related to the damage to the radiator 313. In analyzing the degree of damage to the radiator 313, the first characteristic quantity is the sprung acceleration of the vehicle 10. The second characteristic quantity is the temperature of the coolant. The processing circuit 510 divides the sprung acceleration of the vehicle 10 into multiple values ​​according to the temperature of the coolant, and, similarly to the first embodiment, calculates the frequency distribution of the original data and the extracted data obtained by combining all data extracted through multiple time windows. Figure 55 The frequency distribution of sprung acceleration of vehicle 10 in the raw data is shown when the temperature of the coolant in vehicle 10, which is the subject of analysis, is lower than a predetermined temperature. Figure 56 The frequency distribution of sprung acceleration of vehicle 10 is shown in the raw data when the temperature of the coolant in vehicle 10, the object of analysis, is above a predetermined temperature. For example... Figure 55 and Figure 56As shown, in these frequency distributions, the minimum level is set to zero sprung acceleration, and the sprung acceleration is divided into m levels from "1" to "m". The frequency distribution of the extracted data is also calculated based on the levels corresponding to the original data. In this example, the processing circuit 510 divides the coolant temperature into two categories based on the sprung acceleration of the vehicle 10 contained in the original data and the extracted data: a category below a predetermined temperature and a category above a predetermined temperature. The processing circuit 510 calculates the frequency distribution as described above for each of the two categories of coolant temperature.

[0406] Next, the processing circuit 510 in Figure 4 In the processing of step S140 shown, similarly to the first embodiment, for each of the multiple partitions based on the second feature, the error between the frequency distribution of the first feature in the original data and the frequency distribution of the first feature in the extracted data is calculated. Regarding the error, for example, similar to the first embodiment, the formula for calculating the mean absolute error (MAE), i.e., Formula 1, can be used. In this case, if... Figure 55 and Figure 56 In the example shown, "n" is "m". Similarly, "i" is the number from "1" to "m". If the error is calculated with respect to the entire division using Formula 1 above, then the processing circuit 510 advances the processing to step S150.

[0407] The processing in step S150 is the same as that performed in the first embodiment. The processing circuit 510 uses the error of each segment to determine whether the original data and the extracted data are similar. Then, the processing circuit 510 repeatedly performs steps S110 to S150 by changing the settings of multiple time windows until it can extract data with an error below a threshold. As a result, the storage device 520 stores the cutting patterns where each error is below the threshold. In this way, the processing circuit 510 obtains a cutting pattern for extracted data similar to the original data.

[0408] <Calculation of fatigue damage and notification of failure prediction>

[0409] In step S160, the processing circuit 510 extracts data from the original data based on a data extraction pattern similar to the original data stored in the storage device 520 after processing prior to step S150.

[0410] In the seventh embodiment, similar to the first embodiment, fatigue damage is calculated based on the extracted data as an index value to represent the degree of damage accumulated at the radiator 313.

[0411] Damage accumulates at the radiator 313 due to vibrations generated at the vehicle 10. The greater the sprung acceleration generated at the vehicle 10, the greater the vibration at the vehicle 10, and the greater the accumulated damage at the radiator 313. Damage also accumulates in the radiator 313 due to the inflow of high-temperature coolant. If the temperature of the coolant flowing into the radiator 313 is high, the thermal stress applied to the various structural components of the radiator 313 increases, making it easier for damage to accumulate. In other words, the higher the temperature of the coolant, the greater the accumulated damage at the radiator 313.

[0412] As an example, the processing circuit 510 calculates the fatigue damage of the heat sink 313 using the following method.

[0413] The processing circuit 510 calculates the frequency distribution for each division of the sprung acceleration of the vehicle 10, based on data obtained by dividing the data according to the current coolant temperature. One of the calculated divisions is designated as the reference division. Next, for the data in divisions other than the reference division, the sprung acceleration data included in those divisions is corrected according to the coolant temperature. Regarding this correction, any method that can reflect the deterioration of each structural component caused by thermal stress based on the coolant temperature can be used, taking into account the material and heat resistance temperature of each structural component of the radiator 313. As an example, if the corrected sprung acceleration Ac is calculated by defining a formula incorporating the coolant temperature, the subsequent processing is as follows.

[0414] Based on the corrected sprung acceleration Ac obtained using the formula described above, the processing circuit 510 aggregates all the divided frequency distributions into a single frequency distribution determined as a reference division, and calculates a new frequency distribution in the extracted data, namely the corrected frequency distribution. Next, the processing circuit 510 calculates the fatigue damage degree based on the corrected frequency distribution.

[0415] exist Figure 57The diagram shows an example of a corrected frequency distribution in the analysis of the degree of damage to the heat sink 313. In this corrected frequency distribution, the corrected spring acceleration Ac is classified into six levels, A through F. The frequency Hij for each level represents the number of data points where the corrected spring acceleration Ac is Ai or higher and lower than Aj. For example, data points where the corrected spring acceleration Ac is A2 or higher and lower than A3 are classified into level B, and their frequency is represented as H23. An upper limit frequency Gij is determined for each level. The upper limit frequency Gij represents the maximum number of accumulated damages that would result in fatigue failure at the heat sink 313 if damage caused by the corrected spring acceleration Ac included in the corresponding level is accumulated. As an example, if G34 is L times, fatigue failure will occur at the heat sink 313 if L times of spring acceleration such as that included in the range where the corrected spring acceleration Ac is A3 or higher and lower than A4 occurs. Processing circuit 510 uses the frequency Hij for each level and the upper limit frequency Gij for each level in the corrected frequency distribution to calculate the fatigue damage degree according to Formula 3, similar to the first embodiment. If the processing circuit 510 calculates the fatigue damage degree, it advances the processing to the next step. Figure 4 The step S170 shown.

[0416] Regarding the subsequent steps S170 to S190, the processing circuit 510 performs the same processing as in the first embodiment.

[0417] <Effect of the 7th Embodiment>

[0418] The data center 500 of the information processing device in the seventh embodiment extracts a portion of the raw data collected over a predetermined period using multiple sensors mounted on the vehicle 10, and analyzes the extent of damage accumulated at the radiator 313.

[0419] Data center 500 includes a processing circuit 510 that performs processing according to a program. The raw data includes the sprung acceleration of the vehicle 10 equipped with a radiator 313 as a first characteristic. The raw data includes the temperature of the cooling water flowing into the radiator 313 as a second characteristic. In this data center 500, the processing circuit 510 performs a search process. The search process includes a first process (step S130), which divides the data of the first characteristic into multiple parts according to the second characteristic included in the raw data, and calculates the frequency distribution of the raw data for the first characteristic for each part. The search process includes a second process (step S110), which sets multiple time windows to extract data from a portion of the raw data, such that the sum of the periods of all time windows is shorter than the overall period of the raw data. The search process includes a third process (step S120) to extract data from the raw data through the multiple time windows. The data obtained by combining all the data extracted through the multiple time windows is the extracted data. The search process includes a fourth process (step S130), which divides the data of the first feature quantity into multiple partitions corresponding to the second feature quantity of the original data, and calculates the frequency distribution of the extracted data with respect to the first feature quantity for each partition. The search process includes a fifth process (steps S140 and S150) which calculates the error between the frequency distribution of the original data and the frequency distribution of the extracted data to determine whether the original data and the extracted data are similar. After executing the first process, the processing circuit 510 performs a search process that repeatedly executes the second to fifth processes by changing the settings of multiple time windows. Then, the processing circuit 510 extracts the extracted data with an error below a threshold. Using the extracted data with an error below the threshold, the processing circuit 510 calculates the fatigue damage degree as an indicator value of damage (step S160).

[0420] According to the data center 500, the distribution of characteristic quantities related to the damage to the heat sink 313 can be analyzed using extracted data similar to the original data. Therefore, the data center 500 can obtain analysis results that are close to those obtained using the original data for damage analysis.

[0421] The extracted data obtained through Data Center 500 is obtained by cutting a portion of the original data. Therefore, the amount of extracted data is smaller compared to the original data. The more data used in parsing, the longer the processing time required for parsing the degree of corruption. By using extracted data, Data Center 500 is able to shorten the parsing time compared to using the original data.

[0422] <Effects of the 7th Embodiment>

[0423] In addition to the effects of the first embodiment (1-1) to (1-5), the seventh embodiment also has the following effects.

[0424] (7-1) Data center 500 analyzes the extent of damage to radiator 313. Data center 500's processing circuit 510 sets the sprung acceleration of vehicle 10 equipped with radiator 313 as the first characteristic quantity and the temperature of cooling water flowing into radiator 313 as the second characteristic quantity.

[0425] Damage accumulates at the radiator 313 due to vibrations generated at the vehicle 10. The greater the magnitude of the spring acceleration generated at the vehicle 10, the greater the vibration generated at the vehicle 10, and the greater the accumulated damage at the radiator 313.

[0426] At radiator 313, damage accumulates due to the inflow of high-temperature cooling water. If the temperature of the cooling water flowing into radiator 313 is high, the thermal stress applied to the structural components of radiator 313 increases, making it more prone to cumulative damage. In other words, the higher the temperature of the cooling water, the greater the accumulated damage at radiator 313. Data center 500 uses the above two physical quantities that affect the magnitude of accumulated damage at radiator 313 as feature quantities to obtain extracted data.

[0427] According to data center 500, it is possible to extract data suitable for analyzing the extent of damage to heat sink 313.

[0428] <Amendment to Implementation #7>

[0429] The seventh embodiment described above can be implemented with modifications as described below. The seventh embodiment and the following modifications to the seventh embodiment can be combined with each other without creating technical contradictions.

[0430] The aforementioned data center 500 divides the sprung acceleration of the vehicle 10 into multiple data points according to the temperature of the cooling water to extract data. The data center 500 can use the rotational speed of the pump 314, which circulates the cooling water to the radiator 313, as the first characteristic quantity instead of the sprung acceleration of the vehicle 10.

[0431] The processing circuit 510 of the data center 500 divides the rotational speed of the water pump 314 into multiple segments according to the temperature of the cooling water, and calculates the frequency distribution of the raw and extracted data. For example, Figure 58 The frequency distribution of the rotational speed of water pump 314 is shown in the raw data when the temperature of the cooling water in vehicle 10, which is the object of analysis, is lower than a predetermined temperature. Figure 59 The frequency distribution of the rotational speed of the water pump 314 in the raw data when the temperature of the coolant in the vehicle 10 being analyzed is above a predetermined temperature is shown. In these frequency distributions, the rotational speed of the water pump 314 is set to zero as the smallest level, and the rotational speed of the water pump 314 is divided into m levels from "1" to "m".

[0432] Processing circuit 510 calculates for each division of cooling water temperature. Figure 58 and Figure 59 The frequency distribution of the rotational speed of water pump 314 in the raw and extracted data is shown. Data center 500 is able to extract data from the raw data using the frequency distribution calculated according to the temperature of the cooling water.

[0433] Subsequently, the processing circuit 510, similarly to the above, corrects the rotational speed data of the water pump 314 included in the temperature division of the cooling water, and calculates the corrected rotational speed of the water pump 314. Based on the corrected rotational speed of the water pump 314, the processing circuit 510... Figure 60 As shown, all the frequency distributions from the divisions are aggregated into a single frequency distribution determined by using the division as a baseline, and a new frequency distribution, i.e., the corrected frequency distribution, is calculated from the extracted data. The processing circuit 510, similarly to the above, is based on... Figure 60 The fatigue damage degree of radiator 313 is calculated based on the corrected frequency distribution shown.

[0434] The higher the pressure of the cooling water circulating inside the radiator 313, the more easily the radiator 313 is damaged. When the rotational speed of the water pump 314 is high, the pressure generated by the cooling water circulating inside the radiator 313 becomes higher. That is, the higher the rotational speed of the water pump 314, the more easily the radiator 313 is damaged cumulatively.

[0435] At radiator 313, damage accumulates due to the inflow of high-temperature cooling water. If the temperature of the cooling water flowing into radiator 313 is high, the thermal stress applied to the structural components of radiator 313 increases, making it more prone to cumulative damage. In other words, the higher the temperature of the cooling water, the greater the accumulated damage at radiator 313. Data center 500 uses the above two physical quantities that affect the magnitude of accumulated damage at radiator 313 as feature quantities to obtain extracted data.

[0436] According to data center 500, it is possible to extract data suitable for analyzing the extent of damage to heat sink 313.

[0437] The aforementioned data center 500 divides the sprung acceleration of the vehicle 10 into multiple data points based on the temperature of the cooling water to extract data. The data center 500 can use the internal combustion engine speed (crankshaft speed) of the engine 21 connected to the radiator 313 as the second characteristic quantity, instead of the cooling water temperature.

[0438] The processing circuit 510 of the data center 500 divides the sprung acceleration into multiple values ​​according to the crankshaft rotation speed to calculate the frequency distribution of the raw and extracted data. For example, Figure 61The frequency distribution of sprung acceleration of vehicle 10 is shown in the raw data when the crankshaft speed is outside a given range. Figure 62 The frequency distribution of sprung acceleration of vehicle 10 in the raw data when the crankshaft speed is within a given range is shown. In these frequency distributions, sprung acceleration zero is set as the smallest level, and sprung acceleration is divided into m levels from "1" to "m".

[0439] Processing circuit 510 calculates for each division of crankshaft speed. Figure 61 and Figure 62 The image shows the frequency distribution of sprung accelerations in the raw and extracted data. Data Center 500 can extract extracted data from the raw data using a frequency distribution calculated based on crankshaft rotation speed.

[0440] The aforementioned predetermined range is, for example, a crankshaft speed of RVA or higher but lower than RVB. This range is, for example, a range of crankshaft speeds where the vibration generated at engine 21 coincides with the resonant frequency of vehicle 10. In this case, radiator 313 functions as a dynamic damper that absorbs the vibration of engine 21 to suppress the vibration of vehicle 10. That is, with the same sprung acceleration of vehicle 10, the vibration of radiator 313 is greater when the crankshaft speed is within the predetermined range compared to when the crankshaft speed is outside the predetermined range.

[0441] In the calculation of fatigue damage, the processing circuit 510 calculates the frequency distribution of sprung acceleration in the extracted data for each division of crankshaft rotation speed. At this time, the processing circuit 510... Figure 63 As shown, the sprung acceleration is weighted according to the crankshaft speed. When the crankshaft speed is above RVA and below RVA (i.e., within a predetermined range), the processing circuit 510 multiplies the sprung acceleration data by a weighting coefficient "CV2". When the crankshaft speed is below RVA or above RVA (i.e., outside the predetermined range), the processing circuit 510 multiplies the sprung acceleration data by a weighting coefficient "CV1". The value of "CV2" is set to be greater than the value of "CV1". In other words, the processing circuit 510 corrects the sprung acceleration data in a way that the sprung acceleration is greater when the crankshaft speed is within the predetermined range compared to when it is outside the predetermined range.

[0442] Thus, for each division of crankshaft rotation speed, processing circuit 510 calculates the frequency distribution of sprung acceleration obtained by weighting and correcting the sprung acceleration data. Subsequently, processing circuit 510... Figure 64 As shown, the frequency distributions of each corrected partition are aggregated into a single frequency distribution determined as a baseline partition, and a new frequency distribution, i.e., the corrected frequency distribution, is calculated from the extracted data. The processing circuit 510, similarly to the above, is based on... Figure 64 The corrected frequency distribution is shown. The fatigue damage degree of the heat sink 313 is calculated.

[0443] Damage accumulates at the radiator 313 due to vibrations generated at the vehicle 10. The greater the magnitude of the spring acceleration generated at the vehicle 10, the greater the vibration generated at the vehicle 10, and the greater the accumulated damage at the radiator 313.

[0444] Vibration occurs at engine 21 due to the movement of internal structural components. The magnitude of this vibration varies with crankshaft rotation speed. The vibration at vehicle 10 caused by engine 21 damages radiator 313. In other words, the extent of damage to radiator 313 is affected by crankshaft rotation speed. For example, radiator 313, which functions as a dynamic damper, is prone to significant vibration when crankshaft rotation speed is close to the resonant frequency of vehicle 10. Therefore, radiator 313 is susceptible to significant damage when crankshaft rotation speed is close to its resonant frequency.

[0445] The aforementioned information processing device uses the two physical quantities that affect the extent of damage accumulated at the heat sink 313 as feature quantities to obtain extracted data. According to the data center 500, data suitable for analyzing the degree of damage to the heat sink 313 can be extracted.

[0446] The extent of damage to heat exchangers analyzed by the Data Center 500 is not limited to radiator 313. The Data Center 500 can also analyze the extent of damage to heat exchangers that cool lubricating oil, i.e., oil coolers. In this case, the refrigerant flowing into the heat exchanger is oil. The oil, for example, is ATF. In the above case, the pump that circulates the refrigerant to the heat exchanger is an oil pump.

[0447] <Eighth Implementation>

[0448] Next, refer to Figures 65-71 This section describes an eighth embodiment of the information processing device. Furthermore, the eighth embodiment is an information processing device that analyzes the degree of damage to the engine 21 using the amount of deposits in the intake system of the engine 21 installed in the vehicle 10 as an indicator. The eighth embodiment differs from the first embodiment in the equipment or components used to analyze the degree of damage. In the following description, the parts that differ from the first embodiment will be mainly explained. Detailed descriptions of components that are repeated in the first embodiment are omitted. In the eighth embodiment, the information processing device for analyzing the degree of damage is also the processing circuit 510 of the data center 500.

[0449] <Structure of Engine 21>

[0450] like Figure 65As shown, engine 21 is an internal combustion engine with multiple cylinders 332. Engine 21 is, for example, a gasoline engine. Each cylinder 332 constitutes a combustion chamber for burning a mixture of gasoline and intake air.

[0451] The engine 21 includes a crankcase 331, cylinders 332, a cylinder head 333, an intake passage 342, and an exhaust passage 343. Inside each cylinder 332, a piston 334 and a connecting rod 335 are housed. The connecting rod 335 is connected to a crankshaft 336 housed in the crankcase 331.

[0452] A cylinder head 333 is installed on the upper part of each cylinder 332. The combustion chamber in each cylinder 332 is formed by the cylinder 332 and the cylinder head 333. The cylinder head 333 is equipped with an intake valve 337, an exhaust valve 338, and an ignition device 339.

[0453] Engine 21 is a gasoline engine that supplies gasoline using both port injection and in-cylinder injection. A port injector 340 for port injection is provided in the intake passage 342. A direct injection injector 341 for in-cylinder injection is provided at the cylinder head 333. The ignition device 339, port injector 340, and direct injection injector 341 are each connected to a control device (not shown). The control device is, for example, an ECU that controls fuel injection and ignition in engine 21.

[0454] The intake passage 342 and exhaust passage 343, which connect to each combustion chamber, are respectively connected to the cylinder head 333.

[0455] The intake passage 342 is a passage that introduces intake gas from the outside into each combustion chamber. The downstream end of the intake passage 342 connects to each combustion chamber. The downstream end of the intake passage 342, located within the cylinder head 333, is an intake port. An intake valve 337 is provided at this end. The opening of the intake passage 342 into the combustion chamber is opened and closed by the intake valve 337. The intake system of the engine 21 includes the aforementioned intake passage 342, intake port, and intake valve 337 as structural components.

[0456] The exhaust passage 343 is a passage that guides the exhaust gas from each cylinder 332 into the exhaust system components. The upstream end of the exhaust passage 343 connects to each combustion chamber. The upstream end of the exhaust passage 343, located within the cylinder head 333, is an exhaust port. An exhaust valve 338 is provided at this end. The opening of the exhaust passage 343 into the combustion chamber is opened and closed by the exhaust valve 338.

[0457] <Deposits in the intake system of engine 21>

[0458] After the air-fuel mixture is burned in the combustion chambers of engine 21, the combustion gases are discharged through exhaust passage 343. The combustion gases contain particulate matter produced by combustion. These particulates are, for example, soot containing carbon residue from combustion. If the combustion gases containing these particulates are blown back into the intake system of engine 21, the particulates adhere to the intake system and deposit. These particulates deposited in the intake system are called deposits. For example, deposits are deposited inside the intake port of engine 21. For example, deposits are deposited inside the intake passage 342 of engine 21.

[0459] The deposits in the intake system are affected by the valve overlap in engine 21. Valve overlap is the length of the period during which both intake valve 337 and exhaust valve 338 are open during the combustion cycle in each cylinder 332. The longer the period during which both intake valve 337 and exhaust valve 338 are open, the greater the valve overlap. If the valve overlap is large, more combustion gases are blown back into the intake system, so particles contained in the combustion gases are more likely to be deposited as deposits in the intake system.

[0460] <Data Extraction>

[0461] The information processing terminal 600 sends an indication to the data center 500, using the amount of deposits in the intake system of the engine 21 as an indicator to analyze the degree of damage to the engine 21. Then, similarly to the first embodiment, the data center 500 performs data processing by extracting data from the raw data through multiple time windows.

[0462] Figure 66 Raw data showing characteristic quantities related to the amount of deposits in the intake system of engine 21 of a particular vehicle 10. Figure 66 The raw data shown is a portion of the data equivalent to 100,000 hours from vehicle 10, which is the object of analysis. Figure 66 The raw data shown includes the valve overlap of engine 21 as a feature. The raw data also includes road surface information segmentation based on the location information of the vehicle 10 equipped with engine 21 as a feature. These features are physical quantities correlated with the amount of deposits in the intake system of engine 21.

[0463] Figure 66 (a) shows the valve overlap. Figure 66 (b) shows the road surface information segmentation. Road surface information segmentation is a parameter representing the condition of the road surface on which vehicle 10 travels. The road surface information segmentation includes, for example, paved roads, gravel roads, and dirt roads.

[0464] Data center 500 identifies a segmentation pattern for extracting features from the overall raw data containing the aforementioned feature quantities. Using the extracted data extracted based on the segmentation pattern identified by data center 500, processing circuit 510 analyzes the amount of deposits in the intake system of engine 21 of vehicle 10, which is the object of analysis.

[0465] <Search Processing for Slicing Patterns>

[0466] like Figure 4 As shown, the processing circuit 510 performs the same series of processes as in the first embodiment according to the program.

[0467] In step S100, the processing circuit 510 acquires raw data for a specific vehicle 10. The raw data includes data for analyzing the amount of deposits in the intake system of the engine 21 of the vehicle 10, which is the subject of analysis.

[0468] Next, in the processing of step S110, the processing circuit 510 targets... Figure 66 The raw data of characteristic quantities related to the amount of deposits in the intake system of engine 21, as shown in the first embodiment, are used to determine the cutting mode by setting multiple time windows. The multiple time windows are set in such a way that the sum of the periods of all the time windows is shorter than the period of the entire raw data.

[0469] In step S120, the processing circuit 510, similarly to the first embodiment, extracts data by cutting data through multiple time windows based on the cutting mode determined in step S110.

[0470] In step S130, the processing circuit 510 calculates the frequency distribution of characteristic quantities related to the amount of deposits in the intake system of the engine 21. In the analysis of the amount of deposits in the intake system of the engine 21, the first characteristic quantity is the valve overlap of the engine 21. The second characteristic quantity is the road surface information segmentation assigned based on the position information of the vehicle 10.

[0471] The processing circuit 510 divides the valve overlap into multiple categories according to the road surface information. Similar to the first embodiment, it calculates the frequency distribution of the original data and the extracted data obtained by combining all the data extracted through multiple time windows.

[0472] Figure 67 The frequency distribution of valve overlap is shown in the original data when the road surface information of vehicle 10, which is the object of analysis, is divided into paved roads. Figure 68 The frequency distribution of valve overlap is shown in the raw data when the road surface information of vehicle 10, which is the object of analysis, is a gravel road. Figure 69The frequency distribution of valve overlap is shown in the raw data when the road surface information of vehicle 10, which is the object of analysis, is a dirt road.

[0473] like Figures 67-69 As shown, in these frequency distributions, zero valve overlap is set as the smallest level, and the valve overlap is divided into m levels from "1" to "m". The frequency distribution of the extracted data is also calculated based on the levels corresponding to the original data. In this example, the processing circuit 510 divides the valve overlap contained in the original data and the extracted data into three categories. Each category is a road surface information category divided into paved roads, gravel roads, and dirt roads. The processing circuit 510 calculates the frequency distribution as described above for each of the three road surface information categories.

[0474] Next, the processing circuit 510 in Figure 4 In the processing of step S140 shown, similarly to the first embodiment, for each of the multiple partitions based on the second feature, the error between the frequency distribution of the first feature in the original data and the frequency distribution of the first feature in the extracted data is calculated. Regarding the error, for example, similar to the first embodiment, the formula for calculating the mean absolute error (MAE), i.e., Formula 1, can be used. In this case, if... Figures 67-69 In the example shown, "n" is "m". Similarly, "i" is the number from "1" to "m". If the error is calculated with respect to the entire division using Formula 1 above, then the processing circuit 510 advances the processing to step S150.

[0475] The processing in step S150 is the same as that performed in the first embodiment. The processing circuit 510 uses the error of each segment to determine whether the original data and the extracted data are similar. Then, the processing circuit 510 repeatedly performs steps S110 to S150 by changing the settings of multiple time windows until it can extract data with an error below a threshold. As a result, the storage device 520 stores the cutting patterns where each error is below the threshold. In this way, the processing circuit 510 obtains a cutting pattern for extracted data similar to the original data.

[0476] <Calculation of Sediment Deposition and Notification of Failure Prediction>

[0477] In step S160, the processing circuit 510 extracts data from the original data based on a data extraction pattern similar to the original data stored in the storage device 520 after processing prior to step S150.

[0478] In the eighth embodiment, based on the extracted data, the amount of deposits in the intake system of engine 21 is calculated as an indicator of the degree of damage accumulated at engine 21. The amount of deposits is, for example, the mass of soot deposited in the intake passage 342 of engine 21. The amount of deposits is, for example, the mass of soot deposited at the intake port of engine 21.

[0479] By blowing combustion gases back into the intake system, deposits are deposited in the intake system. The greater the overlap between the intake valve 337 and the exhaust valve 338 when they are simultaneously open, the greater the backflow into the intake system. That is, the greater the valve overlap, the easier it is for deposits to accumulate in the intake system of engine 21.

[0480] The amount of deposits in the intake system of engine 21 changes depending on the road conditions on which vehicle 10 is traveling. Road surface information is determined based on location information as a parameter reflecting road surface conditions. For example, road surface information is a parameter indicating whether the road surface vehicle 10 is traveling on is paved, gravel, or dirt. For instance, when vehicle 10 is traveling on gravel, sand and dust from the road surface easily flow into the intake system. When the particles contained in the sand and dust blow combustion gases back into the intake system, they, along with soot contained in the combustion gases, are deposited as sediment in the intake system. When vehicle 10 is traveling on roads such as gravel or dirt, where particles easily flow into the intake gases, sediment easily accumulates in the intake system.

[0481] As an example, the processing circuit 510 calculates the amount of deposits deposited in the engine 21 using the following method.

[0482] First, for each segment of the road surface information, the processing circuit 510 converts the valve overlap of the engine 21 in the extracted data into the amount of sediment.

[0483] Figure 70 This is a graph showing the relationship between valve overlap and deposit amount in engine 21. The relationship between valve overlap and deposit amount is set differently for each road surface segmentation. Specifically, with the same valve overlap, the deposit amount is set larger when the road surface segmentation is gravel compared to when it is paved. Similarly, with the same valve overlap, the deposit amount is set larger when the road surface segmentation is dirt compared to when it is gravel. That is, for the same valve overlap, the deposit amount increases sequentially from paved to gravel to dirt. The processing circuit 510 is based on... Figure 70 The relationship shown in the graph converts the valve overlap amount included in the extracted data into sediment amount. Therefore, for each segment of the road surface information, the following calculation is performed. Figure 71The frequency distribution of sediment deposition amounts is shown.

[0484] Processing circuit 510 calculates for each segment of road surface information. Figure 71 The frequency distribution of sediment deposition shown is used to calculate the total sediment deposition Ds for each segment of the road surface information according to Formula 7 below.

[0485] [Formula 7]

[0486]

[0487] In Formula 7 above, "n" is the total number of ranks in the frequency distribution. For example, if it is... Figure 71 In the example shown, "n" is "k". "i" is the number that determines the rank in the frequency distribution. For example, if it is Figure 71 In the example shown, "i" is the number from "1" to "k". "di" is the amount of sediment deposited at each frequency in the i-th level. "ti" is the frequency in the i-th level.

[0488] If the processing circuit 510 calculates the total sediment deposition Ds of each segment of the road surface information, then it calculates the total sediment deposition Dp in the original data as a whole according to the following formula 8.

[0489] [Formula 8]

[0490]

[0491] In Formula 8 above, "Lall" represents the period during which the original data was acquired. If it is... Figure 71 In the example shown, "Lall" is 100,000 hours. "Lcut" is the period during which the data was retrieved. If it is... Figure 71 In the example shown, "Lcut" is 20,000 hours. "Dsp" is the total sediment deposition Ds when the road surface information is classified as paved road. "Dsg" is the total sediment deposition Ds when the road surface information is classified as gravel road. "Dsd" is the total sediment deposition Ds when the road surface information is classified as dirt road.

[0492] If the processing circuit 510 calculates the total sediment deposition Dp in the original data set, then it advances the processing to the next step. Figure 4 The step S170 shown.

[0493] In the eighth embodiment, during the processing in step S170, the processing circuit 510 determines whether the total sediment deposition amount Dp is above a boundary value. The boundary value is, for example, the quality of sediment used to predict the increased likelihood of undesirable conditions occurring at the engine 21 based on the total sediment deposition amount Dp being above the boundary value. The processing circuit 510 can predict the increased likelihood of undesirable conditions occurring at the engine 21 based on the total sediment deposition amount Dp reaching the boundary value. Furthermore, undesirable conditions in the engine 21 refer to, for example, knocking.

[0494] If, during the processing in step S170, it is determined that the total sediment deposition amount Dp is above a boundary value (step S170: Yes), the processing circuit 510 advances the processing to step S180. Similarly to the first embodiment, during the processing in step S180, the processing circuit 510 outputs the total sediment deposition amount Dp and a fault prediction.

[0495] If, during the processing in step S170, it is determined that the total sediment deposition amount Dp is lower than the boundary value (step S170: No), the processing circuit 510 advances the processing to step S190. Similarly to the first embodiment, during the processing in step S190, the processing circuit 510 outputs the total sediment deposition amount Dp.

[0496] If step S180 or step S190 is performed, the processing circuit 510 terminates the above series of program-based processes.

[0497] <Effect of the 8th Embodiment>

[0498] The data center 500, as an information processing device in the eighth embodiment, extracts a portion of the raw data collected within a predetermined period using multiple sensors mounted on the vehicle 10, and analyzes the extent of damage accumulated at the engine 21. The data center 500 uses the amount of deposits in the intake system of the engine 21 as an indicator to analyze the extent of damage accumulated at the engine 21.

[0499] Data center 500 includes processing circuit 510 that performs processing according to a program. The raw data includes the valve overlap of engine 21 as a first feature. The raw data includes road surface information based on the location information of vehicle 10 equipped with engine 21 as a second feature. In this data center 500, processing circuit 510 performs search processing. The search processing includes a first process (step S130), which divides the data of the first feature into multiple parts according to the second feature included in the raw data, and calculates the frequency distribution of the raw data for the first feature for each part. The search processing includes a second process (step S110), which sets multiple time windows to extract data from a portion of the raw data, such that the sum of the periods of all time windows is shorter than the overall period of the raw data. The search processing includes a third process (step S120) to extract data from the raw data through multiple time windows. The data obtained by combining all the data extracted through the multiple time windows is the extracted data. The search process includes a fourth process (step S130), which divides the data of the first feature quantity into multiple partitions corresponding to the second feature quantity of the original data, and calculates the frequency distribution of the extracted data with respect to the first feature quantity for each partition. The search process includes a fifth process (steps S140 and S150) which calculates the error between the frequency distribution of the original data and the frequency distribution of the extracted data to determine whether the original data and the extracted data are similar. After executing the first process, the processing circuit 510 performs a search process that repeatedly executes the second to fifth processes by changing the settings of multiple time windows. Then, the processing circuit 510 extracts the extracted data with an error below a threshold. Using the extracted data with an error below the threshold, the processing circuit 510 calculates the amount of deposits in the intake system of the engine 21 as an indicator value of damage (step S160).

[0500] According to the data center 500, the distribution of characteristic quantities related to the amount of deposits in the intake system of engine 21 can be analyzed using extracted data similar to the original data. Therefore, the data center 500 can obtain analysis results that are close to those obtained using corrupted analysis performed with the original data.

[0501] The extracted data obtained through Data Center 500 is obtained by cutting a portion of the original data. Therefore, the amount of extracted data is smaller compared to the original data. The more data used in parsing, the longer the processing time required for parsing the degree of corruption. By using extracted data, Data Center 500 is able to shorten the parsing time compared to using the original data.

[0502] <Effects of the 8th Embodiment>

[0503] In addition to the effects of the first embodiment (1-1) to (1-5), the eighth embodiment also has the following effects.

[0504] (8-1) The data center 500 uses the amount of deposits in the intake system of the engine 21 as an indicator to analyze the degree of damage to the engine 21. The processing circuit 510 of the data center 500 sets the valve overlap of the engine 21 as the first feature and sets the road surface information assigned based on the location information of the vehicle 10 equipped with the engine 21 as the second feature.

[0505] By blowing combustion gases back into the intake system, deposits are deposited in the intake system. The greater the overlap between the intake valve 337 and the exhaust valve 338 when they are simultaneously open, the greater the backflow into the intake system. That is, the greater the valve overlap, the easier it is for deposits to accumulate in the intake system of engine 21.

[0506] The amount of deposits in the intake system of engine 21 changes depending on the road conditions on which vehicle 10 travels. Road surface information is determined based on location information as a parameter reflecting road surface conditions. For example, road surface information is a parameter indicating whether the road surface on which vehicle 10 travels is paved, gravel, or dirt. For instance, when vehicle 10 travels on gravel, sand and dust from the road surface easily flow into the intake system. When the particles contained in the sand and dust blow combustion gases back into the intake system, they, along with soot contained in the combustion gases, are deposited as sediment in the intake system. When vehicle 10 travels on roads such as gravel or dirt, where particles easily flow into the intake gases, sediment easily accumulates in the intake system.

[0507] Data center 500 uses the two physical quantities mentioned above that affect the amount of deposits in the intake system of engine 21 as feature quantities to obtain extracted data. Based on data center 500, data suitable for analyzing the degree of damage to engine 21 can be extracted.

[0508] <Example of a modification to the 8th embodiment>

[0509] The eighth embodiment described above can be implemented with modifications as described below. The eighth embodiment and the following modifications to the eighth embodiment can be combined with each other without creating technical inconsistencies.

[0510] Data Center 500 uses the error of each segment of the road surface information in the frequency distribution of valve overlap to determine whether the original data and extracted data are similar. Data Center 500 can also use the error of each segment of the road surface information in the frequency distribution of sediment deposition to determine whether the original data and extracted data are similar.

[0511] In this case, in step S130, the data center 500 calculates the frequency distribution of valve overlap for the original data and the frequency distribution of valve overlap for the extracted data for each segment of the road information. Subsequently, the data center 500... Figure 70 The relationship shown is used to calculate the frequency distribution of sediment deposition in the original data and the frequency distribution of sediment deposition in the extracted data for each segment of the road information. Then, in step S140, the data center 500 calculates the error between the frequency distribution of sediment deposition in the original data and the frequency distribution of sediment deposition in the extracted data for each segment of the road information. The data center 500 can also use the aforementioned error in the frequency distribution of sediment deposition to determine whether the original data and the extracted data are similar.

[0512] In step S160, data center 500 can also calculate an estimated amount of deposits actually deposited in the air intake system of vehicle 10 based on the total deposit amount Dp. For example, processing circuit 510 of data center 500 calculates the estimated deposit amount based on the total deposit amount Dp, using a formula stored in storage device 520. This formula is, for example, an arbitrary function derived from the relationship between the total deposit amount Dp calculated based on extracted data from a preliminary survey of multiple vehicles 10 and the actual amount of deposits deposited in the air intake systems of the multiple vehicles 10. By using this function based on measured data to calculate the estimated deposit amount based on the total deposit amount Dp, the amount of deposits in the air intake system of vehicle 10 can be analyzed more accurately.

[0513] <Example of Change>

[0514] Furthermore, the following elements are common to all the above embodiments. The following modification examples can be combined and implemented without causing technical inconsistencies.

[0515] In the above embodiments, an example of embodying the information processing device as a data center 500 is shown. Then, an example of the data center 500 performing the calculation of the index value is shown. Conversely, the above information processing device can also be embodying as an information processing terminal 600. In this case, the processing circuit 610 of the information processing terminal 600 performs the calculation of the index value. The above information processing device can also be embodying as a control device for the vehicle 10. In this case, the control device for the vehicle 10 can also perform the calculation of the index value. For example, the second control device 92 of the vehicle 10 can also perform the calculation of the index value.

[0516] • When the error of each partition is below the threshold, the data center 500 determines that the original data and the extracted data are similar. Alternatively, the data center 500 may not use all the errors of each partition in a similarity determination. It can determine that the extracted data and the original data are similar by using only one or two partitions that have a significant impact on the degree of damage to the equipment or components, and when the errors of each of these partitions are below the threshold.

[0517] • When the error of each partition is below a threshold, the data center 500 determines that the data is similar. On the other hand, the data center 500 calculates the sum of the errors for each partition, and when the sum of the calculated errors is below a threshold, it can determine that the extracted data is similar to the original data.

[0518] If we compare the total frequencies in the frequency distributions of each partition, the partitions with more cases matching the second feature have a higher total frequency. Therefore, the more cases matching the second feature, the greater the influence of the errors calculated for each partition on similar judgments based on the sum of these errors. Data from partitions with more cases matching the second feature in the original data also tend to have a greater influence in the analysis of the degree of damage. Therefore, in terms of extracting data suitable for analyzing the degree of damage, it is desirable to have smaller errors in the frequency distributions of the original data and the extracted data for partitions with more cases matching the second feature.

[0519] The modified data center 500 uses extracted data whose sum of errors is below a threshold to analyze the degree of corruption. Therefore, it is not easy to use extracted data with large errors in the frequency distribution of partitions that frequently match the second feature quantity during analysis. That is, it is easier to use extracted data with small errors in the frequency distribution of partitions that frequently match the second feature quantity for analysis. According to this data center 500, the fact that the frequency of the second feature quantity matches varies for each partition can be reflected, allowing for a suitable determination of whether the original data and the extracted data are similar.

[0520] Alternatively, Data Center 500 can omit a portion of the error from each partition in the summation calculation. It can also calculate the sum of errors using only one or two partitions that significantly impact the degree of damage to equipment or components. When the sum of errors calculated in this way is below a threshold, it can be determined that the extracted data is similar to the original data.

[0521] The aforementioned data center 500 determines the similarity between the original and extracted data by calculating the error in the frequency distribution. Alternatively, the data center 500 can also determine the similarity between the original and extracted data without calculating the error. For example, statistical methods such as goodness-of-fit tests can be used to determine similarity when the difference between the original and extracted data is not statistically significant.

[0522] The aforementioned data center 500 uses data from the second feature quantity collected when the first feature quantity is acquired for both data extraction and analysis of the degree of corruption. Conversely, the feature quantities used in data extraction and corruption analysis may not necessarily be the same. For example, other feature quantities acquired when the first and second feature quantities are extracted can be used to analyze the degree of corruption.

[0523] The processing circuit 510 of the data center 500 can also interchange the combination of the first feature and the second feature to extract data and analyze the degree of corruption.

[0524] The aforementioned data center has 500 units. Figure 4 In step S110, multiple time windows can be set by considering the proportion of each partition relative to the whole data. For example, multiple time windows can be set in such a way that the proportion of each partition's data relative to the whole original data is equal to the proportion of each partition's data relative to the whole extracted data.

[0525] The processing circuit 510 includes a CPU (Central Processing Unit), RAM (Random Access Memory), and ROM (Read Only Memory). The processing circuit 510 performs software processing. However, this is merely an example. For instance, the processing circuit 510 may also include dedicated hardware circuitry for processing at least a portion of the software processing performed in the above embodiment. The dedicated hardware circuitry is, for example, an ASIC (Application Specific Integrated Circuit). That is, the processing circuit 510 can be any of the structures described in (a) to (c) below. (a) The processing circuit 510 includes a processing device for executing all processing according to a program and a program storage device such as a ROM for storing the program. That is, the processing circuit 510 includes a software execution device. (b) The processing circuit 510 includes a processing device for executing a portion of the processing according to a program and a program storage device. Furthermore, the processing circuit 510 also includes dedicated hardware circuitry for executing the remaining processing. (c) The processing circuit 510 includes dedicated hardware circuitry for executing all processing. Here, there may be multiple software execution devices and / or dedicated hardware circuits. That is, the above-described processing can be executed by a processing circuit that includes at least one of a software execution device and a dedicated hardware circuit. The processing circuit may also include multiple software execution devices and dedicated hardware circuits. The program storage device, i.e., the computer-readable medium, includes all usable media, i.e., storage devices, that can be accessed by a general-purpose or special-purpose computer. The program may also be stored in a computer-readable non-volatile data storage medium such as a CD-ROM and distributed as a program product. The program may also be provided as a downloadable program product by an information service provider connected to a network such as the Internet.

Claims

1. An information processing apparatus configured to analyze the degree of damage to equipment or components mounted on the vehicle by extracting a portion of raw data collected over a predetermined period using multiple sensors mounted on the vehicle, wherein, The information processing device includes a processing circuit. The processing circuit is configured to perform a first process, which designates a physical quantity in the original data that is related to damage to the device or component as a first feature quantity, designates a physical quantity in the original data that is different from the first feature quantity as a second feature quantity, divides the original data into multiple datasets using the second feature quantity, and calculates the frequency distribution of the first feature quantity in each of the multiple datasets obtained by the division. The processing circuit is configured to repeatedly perform the following processing by changing the settings of multiple time windows used to extract a portion of the original data: The second process involves setting the multiple time windows such that the sum of the durations of all time windows is shorter than the predetermined duration. The third process involves extracting data from the original data using the multiple time windows. The fourth process involves dividing the extracted data obtained by combining all the data extracted through the multiple time windows into multiple datasets, corresponding to the division of the original data into multiple datasets, and calculating the frequency distribution of the first feature in each of the multiple datasets obtained by the division. as well as The fifth step involves using the frequency distribution to determine whether the original data and the extracted data are similar. The processing circuit is configured to use the extracted data, which is similar to the original data, to analyze the degree of damage to the device or the component based on the first feature and the second feature.

2. The information processing apparatus according to claim 1, wherein, The processing circuit is configured to, in the fifth process, calculate the error between the frequency distribution of the original data and the frequency distribution of the extracted data for each of the multiple datasets obtained by partitioning, and determine that the original data and the extracted data are similar when the sum of the calculated errors is below a threshold.

3. The information processing apparatus according to claim 1, wherein, The processing circuit is configured as follows: In the fifth process, for each of the multiple datasets obtained by partitioning, the error between the frequency distribution of the original data and the frequency distribution of the extracted data is calculated. When the error of each of the multiple datasets obtained by partitioning is below a threshold, it is determined that the original data is similar to the extracted data.

4. The information processing apparatus according to any one of claims 1 to 3, wherein, The processing circuit is configured to correct the first feature quantity contained in the extracted data according to the division, and to analyze the degree of damage to the device or the component based on the corrected first feature quantity.

5. The information processing apparatus according to any one of claims 1 to 4, wherein, The information processing device is configured to analyze the degree of damage to the parking lock device that prevents rotation of the transmission output shaft. The processing circuit sets the vehicle speed when the transmission is in the parking position as the first characteristic quantity and the vehicle tilt angle as the second characteristic quantity.

6. The information processing apparatus according to any one of claims 1 to 4, wherein, The information processing device is configured to analyze the degree of damage to the rotating machine. The processing circuit sets the angular acceleration of the rotating machine as the first characteristic quantity and the temperature of the rotating machine or the temperature that is related to the temperature of the rotating machine as the second characteristic quantity.

7. The information processing apparatus according to any one of claims 1 to 4, wherein, The information processing device is configured to analyze the degree of damage to the rotating machine. The processing circuit sets the angular acceleration of the rotating machine as the first characteristic quantity and the temperature of the refrigerant cooling the rotating machine as the second characteristic quantity.

8. The information processing apparatus according to any one of claims 1 to 4, wherein, The information processing device is configured to analyze the degree of damage to the sealing member in the sliding contact of the rotating body. The processing circuit sets the rotational speed of the rotating body as the first characteristic quantity and the temperature of the fluid sealed by the sealing member as the second characteristic quantity.

9. The information processing apparatus according to any one of claims 1 to 4, wherein, The information processing device is configured to analyze the degree of damage to the sealing member in the sliding contact of the rotating body. The processing circuit sets the rotational speed of the rotating body as the first characteristic quantity and the external temperature as the second characteristic quantity.

10. The information processing apparatus according to any one of claims 1 to 4, wherein, The information processing device is configured to analyze the degree of damage to the planetary gears. The processing circuit sets the torque input to the planetary gear as the first characteristic quantity and the temperature of the lubricating oil that lubricates the planetary gear as the second characteristic quantity.

11. The information processing apparatus according to any one of claims 1 to 4, wherein, The information processing device is configured to analyze the degree of damage to the planetary gears. The processing circuit sets the torque input to the planetary gear as the first characteristic quantity and the rotational speed of the pump that sprays lubricating oil to lubricate the planetary gear as the second characteristic quantity.

12. The information processing apparatus according to any one of claims 1 to 4, wherein, The information processing device is configured to analyze the degree of damage to the planetary gears. The processing circuit sets the torque input to the planetary gear as the first characteristic quantity and the tilt angle of the vehicle as the second characteristic quantity.

13. The information processing apparatus according to any one of claims 1 to 4, wherein, The information processing device is configured to analyze the degree of damage to the drive shaft. The processing circuit sets the torque input to the drive shaft as the first characteristic quantity and the steering angle of the steering wheel as the second characteristic quantity.

14. The information processing apparatus according to any one of claims 1 to 4, wherein, The information processing device is configured to analyze the degree of damage to the battery that exchanges power with the electric motor. The processing circuit sets the output of the motor as the first characteristic quantity, and selects one from the group consisting of the battery temperature, the battery SOC (state of charge), the upper limit of the battery's charging power, and the upper limit of the battery's discharging power as the second characteristic quantity.

15. The information processing apparatus according to any one of claims 1 to 4, wherein, The information processing device is configured to analyze the extent of damage to the heat exchanger. The processing circuit sets the acceleration of the vehicle equipped with the heat exchanger as the first characteristic quantity and the temperature of the refrigerant flowing into the heat exchanger as the second characteristic quantity.

16. The information processing apparatus according to any one of claims 1 to 4, wherein, The information processing device is configured to analyze the extent of damage to the heat exchanger. The processing circuit sets the acceleration of the vehicle equipped with the heat exchanger as the first characteristic quantity and the rotational speed of the internal combustion engine connected to the heat exchanger as the second characteristic quantity.

17. The information processing apparatus according to any one of claims 1 to 4, wherein, The information processing device is configured to analyze the extent of damage to the heat exchanger. The processing circuit sets the rotational speed of the pump that circulates the refrigerant to the heat exchanger as the first characteristic value, and sets the temperature of the refrigerant flowing into the heat exchanger as the second characteristic value.

18. The information processing apparatus according to any one of claims 1 to 4, wherein, The information processing device is configured to analyze the degree of damage to the internal combustion engine using the amount of deposits in the intake system as an indicator. The processing circuit sets the valve overlap of the internal combustion engine as the first feature quantity and the road surface information assigned based on the position information of the vehicle equipped with the internal combustion engine as the second feature quantity.

19. An information processing method, wherein a processing circuit extracts a portion of data from raw data collected over a predetermined period using multiple sensors mounted on a vehicle to analyze the degree of damage to equipment or components mounted on the vehicle, wherein... The information processing method includes the following steps: The processing circuit performs a first process, in which a physical quantity related to the damage of the device or the component contained in the original data is set as a first feature quantity, a physical quantity in the original data that is different from the first feature quantity is set as a second feature quantity, the original data is divided into multiple datasets using the second feature quantity, and the frequency distribution of the first feature quantity in each of the multiple datasets obtained by division is calculated. The processing circuit modifies the settings of multiple time windows used to extract a portion of the original data to repeatedly perform the following processing: The second process involves setting the multiple time windows such that the sum of the durations of all time windows is shorter than the predetermined period. The third process involves extracting data from the original data using the multiple time windows. The fourth process involves dividing the extracted data obtained by combining all the data extracted through the multiple time windows into multiple datasets, corresponding to the division of the original data into multiple datasets, and calculating the frequency distribution of the first feature in each of the multiple datasets obtained by the division. as well as The fifth step involves using the frequency distribution to determine whether the original data is similar to the extracted data. as well as The processing circuit uses the extracted data, which is similar to the original data, to analyze the degree of damage to the device or the component based on the first feature and the second feature.

20. A program product that enables processing circuitry to extract a portion of raw data collected over a predetermined period using multiple sensors mounted on a vehicle to analyze the extent of damage to equipment or components mounted on the vehicle, wherein... The program product causes the processing circuit to perform a first process, which designates a physical quantity in the original data related to damage to the device or component as a first feature quantity, designates a physical quantity in the original data that differs from the first feature quantity as a second feature quantity, divides the original data into multiple datasets using the second feature quantity, calculates the frequency distribution of the first feature quantity in each of the multiple datasets obtained from the division, and... The processing circuit is configured to repeatedly perform the following processing by changing the settings of multiple time windows used to extract a portion of the original data: The second process involves setting the multiple time windows such that the sum of the durations of all time windows is shorter than the predetermined duration. The third process involves extracting data from the original data using the multiple time windows. The fourth process involves dividing the extracted data obtained by combining all the data extracted through the multiple time windows into multiple datasets, corresponding to the division of the original data into multiple datasets, and calculating the frequency distribution of the first feature in each of the multiple datasets obtained by the division. as well as The fifth step involves using the frequency distribution to determine whether the original data and the extracted data are similar. The processing circuit uses the extracted data, which is similar to the original data, to analyze the degree of damage to the device or the component based on the first feature and the second feature.

Images