Information processing device

Abstract

An information processing device is provided that can extract data suitable for analyzing the lubrication history of a power transmission device mounted on a vehicle from original data. [Solution] This information processing device includes a first step (S110) of setting multiple time windows to extract data from original data using physical quantities related to the lubrication history of a power transmission device as feature quantities, a second step (S120) of extracting data from the original data, a third step (S130) of calculating a frequency distribution of the feature quantities obtained from the original data and a frequency distribution of the feature quantities obtained from the extracted data, and a fourth step (S140 and S150) of determining whether the original data and the extracted data are similar using the frequency distributions. This information processing device repeatedly executes the search process from the first step to the fourth step by changing the settings of the multiple time windows, and extracts extracted data that is similar to the original data.

Description

[Technical Field]

[0001] The present invention relates to an information processing device. [Background technology]

[0002] Patent Document 1 discloses an information processing device that reduces the size of original data for analysis by compressing the original data for analysis. The original data for analysis is data collected over a predetermined period of time using a sensor mounted on a vehicle.

[0003] The information processing device disclosed in Patent Document 1 compresses data by extracting from the original data data acquired when a certain vehicle speed is reached and data acquired at an inflection point in the vehicle speed. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2008-108247 Summary of the Invention [Problem to be solved by the invention]

[0005] Using data acquired by multiple sensors mounted on a vehicle, it is possible to analyze whether the components of a power transmission device mounted on a vehicle have been properly lubricated or not. If data suitable for analyzing such lubrication history can be extracted from the original data, the extracted data can be used to analyze the lubrication history in a shorter time than when using the original data. The above information processing device uses driving data including vehicle speed, location information, and time. The above information processing device associates vehicle speed with driving location to extract a change pattern in vehicle speed and sets an operation schedule for the engine and motor that minimizes fuel consumption. The above information processing device is unable to extract data suitable for analyzing the lubrication history of the components of a power transmission device mounted on a vehicle from the original data. [Means for solving the problem]

[0006] An information processing device for solving the above problem extracts a portion of data from original data collected over a predetermined period using multiple sensors mounted on a vehicle, and extracts data for analyzing the lubrication history of components in a power transmission device mounted on the vehicle. This information processing device includes a processing circuit. The processing circuit performs a first step of setting multiple time windows to extract data for a portion of the original data such that the sum of the periods of all the time windows is shorter than the predetermined period. The processing circuit performs a second step of extracting data from the original data using the multiple time windows. The processing circuit uses, as a feature, the rotational speed of a rotating body that scoops up lubricating oil stored in a case of the power transmission device to lubricate the components. The processing circuit performs a third step of calculating a frequency distribution of the feature obtained from the original data and a frequency distribution of the feature obtained from extracted data that combines all the data extracted using the multiple time windows. The processing circuit performs a fourth step of determining whether the original data and the extracted data are similar using the frequency distribution. The processing circuit repeatedly executes the first to fourth steps by changing the settings of the plurality of time windows. The information processing device extracts the extracted data that is similar to the original data.

[0007] An information processing device for solving the above problem extracts a portion of data from original data collected over a predetermined period using multiple sensors mounted on a vehicle, and extracts data for analyzing the lubrication history of components in a power transmission device mounted on the vehicle. This information processing device includes a processing circuit. The processing circuit performs a first step of setting multiple time windows to extract data for a portion of the original data such that the sum of the periods of all the time windows is shorter than the predetermined period. The processing circuit performs a second step of extracting data from the original data using the multiple time windows. The processing circuit uses the rotational speed of an oil pump that discharges lubricating oil to lubricate the components as a feature. The processing circuit performs a third step of calculating a frequency distribution of the feature obtained from the original data and a frequency distribution of the feature obtained from extracted data that combines all the data extracted using the multiple time windows. The processing circuit performs a fourth step of determining whether the original data and the extracted data are similar using the frequency distribution. The processing circuit repeatedly executes the first to fourth steps by changing the settings of the plurality of time windows. The information processing device extracts the extracted data that is similar to the original data. [Effects of the Invention]

[0008] According to the information processing device, data suitable for analyzing the lubrication history of components in a power transmission device mounted on a vehicle can be extracted from the original data. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 is a schematic diagram showing the relationship between a data center, which is an embodiment of an information processing device, a vehicle, and an information processing terminal. [Figure 2] FIG. 2 is a cross-sectional view of a power split mechanism, which is an example of a power transmission device. [Figure 3]FIG. 3 is a graph showing some of the original data of physical quantities related to the lubrication history of the components in the power split mechanism, where (a) shows the change in rotation speed of the differential ring gear, (b) shows the change in rotation speed of the sun gear, (c) shows the inclination angle of the vehicle, and (d) shows the rotation speed of the second oil pump. [Figure 4] FIG. 4 is a flowchart showing the flow of processing executed by the processing circuit of the data center. [Figure 5] Figure 5 shows the frequency distribution of the rotational speed of the differential ring gear in the original data. [Figure 6] FIG. 6 shows the frequency distribution of the rotation speed of the second oil pump in the original data. DETAILED DESCRIPTION OF THE INVENTION

[0010] First Embodiment Hereinafter, a first embodiment of an information processing device will be described with reference to FIGS. <Configuration of information processing system> 1 shows the configuration of an information processing system. The information processing system includes a data center 500 equipped with an information processing device, an information processing terminal 600, a plurality of vehicles 10, and a communication network 400. The data center 500 is capable of communicating with the plurality of vehicles 10 and the information processing terminal 600 via the communication network 400.

[0011] <Configuration of Data Center 500> As shown in Fig. 1, 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, and includes a CPU that executes processing according to a program, and a ROM in which the program is stored. The storage device 520 is capable of storing large amounts of data. The communication device 530 performs wired or wireless communication via the communication network 400. The communication device 530 includes hardware such as a network adapter, various types of communication software, or a combination of these.

[0012] <Configuration of information processing terminal 600> 1, 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 in which the program is stored. The storage device 620 is capable of storing large amounts of data. The communication device 630 performs wired or wireless communication via the communication network 400. The communication device 630 includes hardware such as a network adapter, various types of communication software, or a combination of these.

[0013] The information processing terminal 600 is, for example, a personal computer. <Vehicle 10 Configuration> The vehicle 10 includes a communication device 90. The communication device 90 transmits data acquired by the vehicle 10 to a data center 500 via a communication network 400. The vehicle 10 includes a hybrid mechanism 20 and a vehicle control unit 80. The hybrid mechanism 20 includes an engine 30, a first motor generator 31, a second motor generator 32, and a power split device 40. The vehicle control unit 80 includes a first control device 81 that controls the engine 30 and a second control device 82 that controls the first motor generator 31 and the second motor generator 32. The vehicle control unit 80 includes a plurality of sensors that collect data. The first control device 81 includes a CPU that controls the operating state of the engine 30. The first control device 81 controls the engine 30 based on data collected by the sensors. The second control device 82 includes a CPU that controls the first motor generator 31 and the second motor generator 32. The second control device 82 controls the first motor generator 31 and the second motor generator 32 based on data collected by the sensors. Examples of data collected by the vehicle control unit 80 include the crankshaft rotation speed, the motor generator rotation speed, the rotation speed of the gears and shafts that make up the power split mechanism 40, and the temperature of the lubricating oil.

[0014] <Configuration of power split mechanism 40> The power split mechanism 40 is a power transmission device that transmits power generated by the engine 30, the first motor generator 31, and the second motor generator 32 to the drive wheels of the vehicle 10. As shown in FIG. 2, the power split mechanism 40 includes a planetary gear unit 42, a reduction gear 49, a motor gear 52, and a differential ring gear 53 inside a case 41. The power split mechanism 40 shown in FIG. 2 is used, for example, in a front-wheel drive vehicle equipped with a hybrid mechanism 20 at the front of the vehicle. The planetary gear unit 42 includes a sun gear 45, three pinion gears 46, a planetary carrier 43, and a ring 44. The sun gear 45 is located at the center of the planetary gear unit 42. The sun gear 45 is connected to the first motor generator 31. The three pinion gears 46 are supported by the planetary carrier 43 and arranged around the sun gear 45. The ring 44 has a ring gear 47 on its inner circumferential surface and an output gear 48 on its outer circumferential surface. The rotation axis of the sun gear 45, the rotation axis of the planetary carrier 43, the rotation axis of the ring 44, and the engine output shaft 33 are all on the same straight line. The engine output shaft 33 is the output shaft of the engine 30. The output of the engine 30 is input to the planetary carrier 43.

[0015] Axis lines S1 to S4 are shown in Figure 2. Axis line S1 is an axis that passes through the rotation axis of sun gear 45, the rotation axis of planetary carrier 43, the rotation axis of ring 44, and engine output shaft 33. Axis line S2 is an axis that passes through the rotation axis of reduction gear 49. Axis line S3 is an axis that passes through the output shaft of second motor generator 32 and the rotation axis of motor gear 52. Motor gear 52 is fixed to the output shaft of second motor generator 32. Axis line S4 is an axis that passes through the rotation axis of differential ring gear 53.

[0016] The output of the engine 30 input to the planetary carrier 43 is distributed from a pinion gear 46 to a sun gear 45 and a ring gear 47. The first motor generator 31 is, for example, a rotating machine used for both power generation and propulsion. The torque distributed to the ring gear 47 rotates the ring 44, thereby rotating an output gear 48. The output gear 48 is meshed with a large reduction gear 50 of a reduction gear 49 arranged on the S2 shaft. Furthermore, the large reduction gear 50 is also meshed with a motor gear 52 provided in the second motor generator 32 arranged on the S3 shaft. The second motor generator 32 is, for example, a rotating machine used for propulsion. As shown in FIG. 2 , the reduction gear 49 includes a small reduction gear 51 in addition to the large reduction gear 50. The small reduction gear 51 is meshed with a differential ring gear 53 arranged on the S4 shaft. The driving force of differential ring gear 53 is transmitted to the drive wheels via the differential device and the drive shaft. With the above configuration, power split mechanism 40 combines the outputs of engine 30 and first motor-generator 31 on S1 shaft and the output of second motor-generator 32 on S3 shaft into one power at reduction gear 49 and transmits it to the drive wheels.

[0017] <Lubrication of Components in Power Split Mechanism 40> The components of the power split mechanism 40 are lubricated by the lubricating oil in the case 41 of the power split mechanism 40. The components are lubricated by the lubricating oil being scooped up by the rotating parts immersed in the lubricating oil, the lubricating oil being splashed from the rotating parts to which the lubricating oil is attached, and the lubricating oil being supplied by the oil pump. The lubricating oil is stored in the lower part of the case 41. The dashed-dotted line in FIG. 2 indicates the liquid level of the lubricating oil stored in the case 41. As shown in FIG. 2, the lower part of the differential ring gear 53 is immersed in the lubricating oil. Therefore, when the differential ring gear 53 rotates, the scooped-up lubricating oil lubricates the gears in the case 41. Meanwhile, the first oil pump 56 and the second oil pump 57 suck up the lubricating oil stored in the lower part of the case 41 and send it to the supply path. In FIG. 2 , the intake ports through which the first oil pump 56 and the second oil pump 57 draw lubricating oil are located on the right side of the partition wall 55 provided on the case 41. The two oil pumps send the lubricating oil drawn from the intake ports to a lubricating oil supply path and distribute it to each component of the power split mechanism 40. For example, the first oil pump 56 is provided as an oil pump connected to the pump drive gear 54 that meshes with the differential ring gear 53. In this case, when the rotation speed of the differential ring gear 53 is high, the distribution of lubricating oil via the first oil pump 56 becomes more intense. In addition, the second oil pump 57 is, for example, an oil pump connected to the engine output shaft 33 of the engine 30. In this case, when the rotation speed of the engine 30 is high, the distribution of lubricating oil via the second oil pump 57 becomes more intense. Furthermore, the rotating components of the power split mechanism 40 that are lubricated by the scooping up or distribution of lubricating oil scatter the lubricating oil adhering to those components into the case 41 as they rotate. Therefore, when the rotation speed of the rotating parts is high, the lubricating oil splashes more violently. Therefore, the lubrication state of each component of power split mechanism 40 is determined by the lubricating oil scooped up by differential ring gear 53, the lubricating oil sprayed by first oil pump 56 and second oil pump 57, and the lubricating oil splashed by each rotating part.

[0018] <Data Extraction> The information processing terminal 600 is used to analyze the lubrication history of components in a power split device 40, which is an example of a power transmission device mounted on a vehicle. Analyzing the lubrication history involves using data acquired by multiple sensors mounted on the vehicle to determine whether the components in the power transmission device mounted on the vehicle have been properly lubricated or not. When analyzing the lubrication history, the information processing terminal 600 transmits instructions to the data center 500. Upon receiving the instructions, the processing circuit 510 of the data center 500 performs analysis using a portion of the vast amount of data stored in the storage device 520 of the data center 500. The data to be used is selected from the vast amount of data stored in the storage device 520 according to the purpose of the analysis. This data includes data on physical quantities related to the lubrication history of the components collected using multiple sensors mounted on the vehicle 10. These physical quantities are referred to as feature quantities. The processing circuit 510 uses the feature quantities to analyze the lubrication history of the components in the power split device 40 of a specific vehicle 10. The number of features used in the analysis may be one or two or more.

[0019] To perform such an analysis, the processing circuitry 510 utilizes a large amount of data collected over a long period of time, and the analysis requires a large amount of calculations, which can take a long time.

[0020] Therefore, it is conceivable to extract extracted data that captures the overall characteristics of the original data from the large amount of data that is the original data. If such extracted data can be extracted, the processing circuit 510 can use the extracted data to perform analysis in a shorter time. For example, when analyzing the lubrication history of the above-mentioned components after 100,000 hours of operation, the processing circuit 510 extracts extracted data for 20,000 hours from the original data for 100,000 hours.

[0021] FIG. 3 shows original data of feature quantities related to the power split mechanism 40. The original data shown in FIG. 3 is a portion of 100,000 hours of data for one vehicle 10. The original data shown in FIG. 3 includes four feature quantities. The first is the rotational speed of a rotor that scoops up lubricating oil stored in the case 41 of the power split mechanism 40 by rotation to lubricate the above-mentioned components. In this embodiment, this rotational speed is the rotational speed of the differential ring gear 53. The second is the rotational speed of a second rotor that, in the case 41 of the power split mechanism 40, scatters adhering lubricating oil by rotation to lubricate the above-mentioned components. In this embodiment, the rotational speed of the second rotor is the rotational speed of the sun gear 45. The third is the inclination angle of the vehicle 10. The fourth is the rotational speed of the second oil pump 57, which is an oil pump that discharges lubricating oil to lubricate the above-mentioned components.

[0022] Fig. 3(a) shows the rotational speed of the differential ring gear 53 among the data for 100,000 hours. Fig. 3(b) shows the rotational speed of the sun gear 45 among the data for 100,000 hours. Fig. 3(c) shows the tilt angle of the vehicle 10. Fig. 3(d) shows the rotational speed of the second oil pump 57.

[0023] The extracted data is created by extracting data from the original data using multiple time windows. In Figure 3, three time windows, the first time window W_1, the second time window W_2, and the third time window W_3, are shown by dashed lines as examples of multiple time windows. The start and end of each time window are set so that they do not overlap. In this example, 20,000 hours of data is extracted as the extracted data. Therefore, the start and end of each time window are set so that the total length of all the time windows is 20,000 hours.

[0024] The data center 500 searches for settings of the start and end of each time window that indicate a cutout pattern for extracting extracted data that captures the characteristics of the entire original data. The data center 500 stores information on the cutout pattern for extracting the extracted data described above in the storage device 520. The stored cutout pattern information is information on the settings of each time window found by the search.

[0025] The processing circuit 510 extracts data from the original data based on the information of the extraction pattern stored in the storage device 520 . <Searching for extraction patterns> 4 is a flowchart showing a series of processes related to the extraction pattern search process. This series of processes is executed by the processing circuit 510 of the data center 500 according to a program. The following shows an example in which the processing circuit 510 uses two feature amounts, a first feature amount and a second feature amount. In this case, the first feature amount is the rotation speed of the differential ring gear 53. The second feature amount is the rotation speed of the sun gear 45.

[0026] 4, the processing circuit 510 acquires original data in the process of step S100. The original data is a portion of data selected from the vast amount of data stored in the storage device 520 of the data center 500 according to the purpose of analysis.

[0027] The original data for analyzing the lubrication history of the components in the power split mechanism 40 of one vehicle 10 is data of the target vehicle 10 selected from a huge amount of data of multiple vehicles 10 .

[0028] Next, in step S110, the processing circuit 510 sets a plurality of time windows to extract extracted data from the original data. In the example shown in Fig. 3, the duration of each time window is the same. As shown in Fig. 3, the data extracted by each extraction window is data of each feature amount for the same period.

[0029] Each time the processing circuit 510 executes the process of step S110, it randomly sets the number of time windows, the start time of each time window, and the end time of each time window. At this time, the processing circuit 510 sets each time window so that the time windows do not overlap. In this way, the processing circuit 510 randomly sets multiple time windows so that the total duration of all the time windows is a predetermined duration. In the process of step S110, the processing circuit 510 may set multiple time windows by fixing the duration of each time window to a constant value, as shown in FIG. 3. In the process of step S110, the processing circuit 510 may set multiple time windows by fixing the number of time windows to a constant value.

[0030] In this way, a cut-out pattern for cutting out data from the original data is determined by setting a plurality of time windows through the process of step S110. After determining the cut-out pattern in this way, processing circuit 510 advances the process to step S120.

[0031] In the process of step S120, the processing circuit 510 extracts data from the original data using the determined extraction pattern. That is, in the process of step S120, the processing circuit 510 extracts data from the original data using the multiple time windows that have been set. Then, the processing circuit 510 combines all of the data extracted using the multiple time windows to create extracted data.

[0032] Next, in the process of step S125, the processing circuit 510 classifies the data of the first feature amount included in the original data into a plurality of categories based on the data of the second feature amount when the first feature amount was collected. Similarly, the processing circuit 510 classifies the data of the first feature amount included in the extracted data into a plurality of categories based on the data of the second feature amount so as to correspond to the categories of the original data.

[0033] In the processing of step S130, the processing circuit 510 calculates the frequency distribution of the first feature amounts of the original data and the first feature amounts of the extracted data for each of the categories based on the data of the first feature amounts of the original data and the extracted data classified into multiple categories.

[0034] A frequency distribution classifies data of the first feature into multiple classes and shows the distribution of the frequency, which is the number of data in each class. Because the total frequency of the first feature contained in the original data and the extracted data differs between them, the frequency distribution of the original data and the extracted data cannot be simply compared. If 20,000 hours of extracted data are extracted from 100,000 hours of original data, the total frequency of the extracted data will be approximately one-fifth of the total frequency of the original data. In this case, multiplying the frequency of each class of the extracted data by five can obtain a frequency distribution of the extracted data with a total frequency equivalent to that of the original data. Even without using the above method, the data distribution in the original data and the data distribution in the extracted data can be compared by calculating a relative frequency distribution as the frequency distribution of the original data and the extracted data. A relative frequency distribution is a frequency distribution that indicates the percentage of the frequency of a class relative to the total frequency.

[0035] FIG. 5 shows a frequency distribution of the rotation speed of the differential ring gear 53, which is the first feature quantity, obtained from the original data of the vehicle 10. In this frequency distribution, the rotation speed is divided into m classes, from "1" to "m," with zero rotation speed being the smallest class. When dividing into multiple classes based on the rotation speed of the sun gear 45, which is the second feature quantity, a frequency distribution of the rotation speed of the differential ring gear 53 in the original data, as shown in FIG. 5, is calculated for each class of the second feature quantity. For example, the data can be divided into a class where the rotation speed of the sun gear 45 is less than a predetermined speed and a class where the rotation speed of the sun gear 45 is equal to or greater than the predetermined speed. In this case, two frequency distributions are calculated: a frequency distribution of the rotation speed of the differential ring gear 53 when the rotation speed of the sun gear 45 is less than the predetermined speed, and a frequency distribution of the rotation speed of the differential ring gear 53 when the rotation speed of the sun gear 45 is equal to or greater than the predetermined speed. Furthermore, in accordance with the above-mentioned divisions and classes, the frequency distribution of the rotation speed of the differential ring gear 53 is calculated for each division of the rotation speed of the sun gear 45 for the extracted data.

[0036] Next, in the process of step S140 shown in Fig. 4, the processing circuit 510 calculates the error between the frequency distribution of the first feature obtained from the original data and the frequency distribution of the first feature obtained from the extracted data for each of a plurality of sections based on the second feature. For example, the processing circuit 510 calculates a mean absolute error (MAE). The mean absolute error (MAE) is expressed by the following equation 1.

[0037]

number

[0038] In the above formula 1, "n" is the total number of classes in the frequency distribution. For example, in the example shown in Figure 5, "n" is "m". "i" is a number that identifies a class in the frequency distribution. For example, in the example shown in Figure 5, "i" is a number from "1" to "m". "Y" is the frequency of the first feature in the corresponding class of the original data. "y" is the frequency of the first feature in the corresponding class of the extracted data.

[0039] As shown in the above formula 1, the processing circuit 510 calculates, for each section, the error as the sum of the errors in the frequencies in each class of the first feature between the frequency distribution obtained from the original data and the frequency distribution obtained from the extracted data.

[0040] After calculating the errors for all the segments, the processing circuit 510 proceeds to step S150. In step S150, the processing circuit 510 determines whether all of the calculated errors for each segment are equal to or less than a threshold value. The threshold value is a value for determining whether extracted data having a frequency distribution similar to that of the original data has been extracted using the set extraction pattern. The magnitude of this threshold value is set in advance so that it can be determined that extracted data having a frequency distribution similar to that of the original data has been extracted based on whether the errors are equal to or less than the threshold value. This threshold value can be set to a different value for each segment.

[0041] In the process of step S150, if it is determined that all of the errors for each section are equal to or less than the threshold value (step S150: YES), processing circuit 510 proceeds to the process of step S160.

[0042] On the other hand, if it is determined in the process of step S150 that any of the errors is greater than the threshold value (step S150: NO), the processing circuit 510 returns the process to step S110. That is, the processing circuit 510 starts the process of setting new multiple time windows in order to reset the time windows and extract extracted data from the original data.

[0043] In this way, processing circuit 510 repeats the processes of steps S110 to S150 using the frequency distribution of the first feature amount divided by the second feature amount until extracted data similar to the original data can be extracted.

[0044] As a result, in step S160, the processing circuit 510 stores in the storage device 520 an extraction pattern in which all of the errors for each section are equal to or less than the threshold. The extracted data extracted using this extraction pattern is data in which the distribution of physical quantities related to the lubrication history is similar to that of the original data. Therefore, the extracted data can be used instead of the original data to analyze the lubrication history of the components of the power split mechanism 40.

[0045] <Operation of the First Embodiment> The data center 500, which is an information processing device of the first embodiment, extracts some data from original data collected over a predetermined period of time using a plurality of sensors mounted on the vehicle 10. This data is extracted data for analyzing the lubrication history of components in the power split mechanism 40, which is a power transmission device mounted on the vehicle 10.

[0046] The data center 500 includes a processing circuit 510. The original data includes, as a first feature, the rotation speed of the differential ring gear 53, which is the rotating body. The original data includes, as a second feature, the rotation speed of the sun gear 45, which is the second rotating body. In the data center 500, the processing circuit 510 executes a search process. The search process includes a first step (step S110) of setting multiple time windows to extract data for a portion of the original data so that the total duration of all the time windows is shorter than the entire duration of the original data. The search process includes a second step (step S120) of extracting data from the original data using the multiple time windows. The extracted data is data obtained by combining all of the data extracted from the original data using the multiple time windows. The search process also includes a third step (steps S125 and S130) of dividing the original data into multiple groups based on the second feature included in the original data and calculating a frequency distribution of the first feature obtained from the original data for each group. On the other hand, the search process also includes a third step (steps S125 and S130) of dividing the extracted data into multiple segments using the second feature amount corresponding to the multiple segments of the original data and calculating a frequency distribution of the first feature amount obtained from the extracted data for each segment. The search process also includes a fourth step (steps S140 and S150) of calculating an error between the frequency distribution of the original data and the frequency distribution of the extracted data for each segment and determining whether the original data and the extracted data are similar. The processing circuit 510 executes the search process by repeatedly executing the processes from steps 1 to 4 while changing multiple time window settings. As a result, the processing circuit 510 finds a data extraction pattern in which all of the errors for each segment are below a threshold. The processing circuit 510 then stores the data extraction pattern when extracted data similar to the original data is extracted in the storage device 520 (step S160).

[0047] In the data center 500, the rotational speed of the differential ring gear 53, which is the rotating body, is used as a first characteristic quantity. The lubricating oil stored in the case 41 of the power split mechanism 40 is scooped up by the rotating body to lubricate the components. The higher the rotational speed of the rotating body, the more vigorously the lubricating oil in the case 41 is scooped up. On the other hand, in the data center 500, the rotational speed of the sun gear 45, which is the second rotating body, is used as a second characteristic quantity. The higher the rotational speed of the sun gear 45, which is the second rotating body, the more vigorously the lubricating oil is scattered from the second rotating body. In other words, the rotational speed of the second rotating body affects the lubrication history of the components.

[0048] The data center 500 determines the similarity between the original data and the extracted data using the frequency distribution of the first feature amount divided into multiple groups according to the second feature amount. That is, the data center 500 extracts data using the two feature amounts that affect the lubrication history. Therefore, when analyzing the lubrication history using the extracted data, it is possible to obtain analysis results that are similar to the results of analyzing the lubrication history using the original data.

[0049] <Effects of the first embodiment> (1-1) According to the data center 500, it is possible to extract, from the original data, extracted data suitable for analyzing the lubrication history of components in a power transmission device mounted on the vehicle 10.

[0050] (1-2) The processing circuit 510 of the data center 500 sets a physical quantity related to the lubrication history of a component, different from the feature quantity included in the original data, as a second feature quantity. In a third step, the processing circuit 510 divides the original data into multiple data sets using the second feature quantity, calculates a frequency distribution of the feature quantity obtained from the original data for each divided data set, and divides the extracted data into multiple data sets using the second feature quantity in correspondence with the multiple data divisions of the original data, and calculates a frequency distribution of the feature quantity obtained from the extracted data for each divided data set. In a fourth step, the processing circuit 510 calculates the error between the frequency distribution of the original data and the frequency distribution of the extracted data for each divided data set. If all of the calculated errors for each division are below a threshold, the processing circuit 510 determines that the original data and the extracted data are similar.

[0051] The lubrication history of components in the power split mechanism 40, which is a power transmission device, is affected by differences in the second feature values. The data center 500 determines, for each classification of the second feature values, whether the error between the frequency distribution of the original data and the frequency distribution of the extracted data is equal to or less than a threshold value. If all of the errors for each classification are equal to or less than the threshold value, the original data and the extracted data are determined to be similar. Therefore, extracted data with small feature error in each classification of the multiple classifications based on the second feature values ​​can be extracted. Therefore, the data center 500 can extract extracted data from the original data while taking into account the influence of the second feature values ​​on the lubrication history.

[0052] (1-3) The processing circuit 510 of the above-mentioned data center 500 determines, as the second characteristic quantity, the rotational speed of a second rotating body that lubricates components by scattering adhering lubricating oil through rotation within the case 41 of the power split mechanism 40, which is a power transmission device.

[0053] The higher the rotational speed of the second rotating body, the more vigorously lubricating oil splashes from the second rotating body. In other words, the rotational speed of the second rotating body affects the lubrication history of the component. The data center 500 determines the similarity between the original data and the extracted data using the rotational speed of the second rotating body as a second feature. Therefore, the data center 500 extracts extracted data whose frequency distribution of the feature, divided into multiple groups according to the rotational speed of the second rotating body, is similar to that of the original data. The data center 500 can extract extracted data from the original data, taking into account the effect of the rotational speed of the second rotating body on the lubrication history.

[0054] <Modification of the first embodiment> The above-described first embodiment can be modified as follows: The first embodiment and the following modifications of the first embodiment can be combined with each other to the extent that they are not technically inconsistent.

[0055] In the data center 500, the second characteristic quantity is the rotational speed of the sun gear 45, which is a second rotating body that lubricates components by scattering adhering lubricating oil as it rotates within the case 41 of the power split mechanism 40. However, the second rotating body is not limited to the sun gear 45. For example, the second rotating body may be the first motor-generator 31 or the planetary carrier 43.

[0056] The data center 500 can also use the tilt angle of the vehicle 10 as the second feature amount instead of the rotation speed of the second rotating body. For example, an example in which the rotation speed of the differential ring gear 53, which is the first feature, is divided into multiple categories based on the inclination angle of the vehicle 10 is shown below. FIG. 5 shows a frequency distribution of the rotation speed of the differential ring gear 53 obtained from the original data. In this frequency distribution, the rotation speed is divided into m categories, from "1" to "m," with zero rotation speed being the smallest category. When dividing into multiple categories based on the inclination angle of the vehicle 10, which is the second feature, a frequency distribution of the rotation speed in the original data, as shown in FIG. 5, is calculated for each category of the second feature. The categories can be, for example, three categories: a category for a positive inclination angle of the vehicle 10, a category for a zero inclination angle of the vehicle 10, and a category for a negative inclination angle of the vehicle 10. When the inclination angle of the vehicle 10 is positive, the vehicle 10 is on an uphill slope. When the inclination angle of the vehicle 10 is negative, the vehicle 10 is on a downhill slope. Then, a frequency distribution of the rotation speed obtained from the original data is calculated for each of these three categories. Furthermore, a similar frequency distribution is calculated for the extracted data corresponding to the above categories and classes. Using the frequency distribution calculated in this way, it is possible to determine the similarity between the original data and the extracted data.

[0057] When the inclination angle of the vehicle 10 changes the level of the lubricating oil stored in the case 41 of the power split mechanism 40, the immersion state of the rotating body, which scoops up the lubricating oil, in the lubricating oil changes. Because the immersion state of the rotating body affects the scooping up of the lubricating oil by the rotating body, the inclination angle of the vehicle 10 affects the lubrication history of the component parts. The data center 500 determines the similarity between the original data and the extracted data using the inclination angle of the vehicle 10 as the second feature. Therefore, the data center 500 extracts extracted data in which the frequency distribution of the first feature, which is divided into multiple groups according to the inclination angle of the vehicle 10, is similar to that of the original data. The data center 500 can extract extracted data from the original data, taking into account the effect of the vehicle inclination angle on the lubrication history.

[0058] Furthermore, the data center 500 may use the temperature of the lubricant that lubricates the components or the rotation speed of the oil pump as the second feature. When the temperature of the lubricant increases, the viscosity of the lubricant decreases, and the supply of the lubricant to the components changes.

[0059] The data center 500 determines that the extracted data is similar to the original data when all of the errors for each category based on the second feature amount are below a threshold. The data center 500 does not need to use all of the errors for each category in the similarity determination. It can use only categories that have a large impact on the lubrication history of one or more of the components, and determine that the extracted data is similar to the original data when all of the errors for those categories are below a threshold.

[0060] Furthermore, the data center 500 calculates the sum of errors for each section based on the second feature amount, and when the sum of errors is equal to or less than a threshold value, it can determine that the extracted data is similar to the original data.

[0061] Second Embodiment Next, a second embodiment will be described with reference to FIGS. 1 to 4 and 6. The second embodiment differs from the first embodiment in that only one feature is used. The feature is the rotational speed of an oil pump that discharges lubricating oil to lubricate the components of the power split mechanism 40. In the following description, differences from the first embodiment will be mainly described. Detailed descriptions of components that overlap with those in the first embodiment will be omitted. In the second embodiment, the information processing device that extracts data is a processing circuit 510 in a data center 500.

[0062] As shown in FIG. 4, the processing circuit 510 executes a series of processes similar to those in the first embodiment. In step S100, the processing circuit 510 acquires original data for a specific vehicle 10. The original data includes a feature, which is the rotational speed of an oil pump that discharges lubricating oil to lubricate the components. The rotational speed of the oil pump is, for example, the rotational speed of the second oil pump 57.

[0063] Next, in the process of step S110, the processing circuit 510 determines an extraction pattern by setting multiple time windows for the original data of the rotation speed of the second oil pump 57, which is the oil pump shown in Fig. 3(d), in the same manner as in the first embodiment. The multiple time windows are set so that the total period of all the time windows is shorter than the period of the entire original data.

[0064] In step S120, the processing circuit 510 creates extracted data by extracting data using a plurality of time windows based on the extraction pattern determined in step S110, as in the first embodiment.

[0065] In the second embodiment, the second feature amount is not used, and therefore the process of step S125 is not executed. In the process of step S130, the processing circuit 510 calculates a frequency distribution of the above-mentioned feature amount. Using the rotation speed of the second oil pump 57 as a feature amount, the processing circuit 510 calculates a frequency distribution of the feature amount obtained from the original data and a frequency distribution of the feature amount obtained from the extracted data. FIG. 6 shows a frequency distribution of the rotation speed of the second oil pump 57 obtained from the original data of the vehicle 10 to be analyzed. In this frequency distribution, the rotation speed is divided into m classes from "1" to "m," with zero rotation speed being the smallest class. The processing circuit 510 also calculates a frequency distribution as shown in FIG. 6 for the extracted data, corresponding to the above classes.

[0066] Next, in the process of step S140 shown in FIG. 4, the processing circuit 510 calculates the error between the frequency distribution of the feature quantities obtained from the original data and the frequency distribution of the feature quantities obtained from the extracted data, as in the first embodiment. The error can be calculated, for example, using Equation 1, which is the same formula for calculating the mean absolute error (MAE) as in the first embodiment. In the example shown in FIG. 6, "n" is "m". Similarly, "i" is a number ranging from "1" to "m". "Y" is the frequency of the feature quantity in the corresponding class of the original data. "y" is the frequency of the feature quantity in the corresponding class of the extracted data. After calculating the error using Equation 1, the processing circuit 510 proceeds to step S150.

[0067] In the process of step S150, the processing circuit 510 determines whether the calculated error is equal to or smaller than a threshold value. The magnitude of the threshold value is set in advance so that, as in the first embodiment, it can be determined that extracted data having a frequency distribution similar to the frequency distribution in the original data has been extracted based on whether the error is equal to or smaller than the threshold value. In the process of step S150, if it is determined that the error is equal to or smaller than the threshold value (step S150: YES), as in the first embodiment, the processing circuit 510 records the extraction pattern in the process of step S160. On the other hand, if it is determined that the error is larger than the threshold value (step S150: NO), as in the first embodiment, the processing circuit 510 returns the process to step S110.

[0068] In this way, processing circuit 510 repeats the processes of steps S110 to S150 by changing the settings of multiple time windows until extracted data in which the error is equal to or less than the threshold is extracted. As a result, as in the first embodiment, in the process of step S160, an extraction pattern in which the error is equal to or less than the threshold is stored in storage device 520.

[0069] <Operation of the Second Embodiment> A data center 500, which is an information processing device of the second embodiment, extracts some data from original data collected over a predetermined period of time using a plurality of sensors mounted on the vehicle 10. This data is extracted data for analyzing the lubrication history of components in the power split mechanism 40, which is a power transmission device mounted on the vehicle 10.

[0070] The data center 500 includes a processing circuit 510. The original data includes the rotation speed of the second oil pump 57 as a feature. In the data center 500, the processing circuit 510 executes a search process. The search process includes a first step (step S110) of setting multiple time windows to extract data for a portion of the original data such that the sum of the periods of all the time windows is shorter than the entire period of the original data. The search process includes a second step (step S120) of extracting data from the original data using the multiple time windows. The extracted data is obtained by combining all of the data extracted from the original data using the multiple time windows. The search process includes a third step (step S130) of calculating a frequency distribution of the feature values ​​obtained from the original data and a frequency distribution of the feature values ​​obtained from the extracted data. The search process includes a fourth step (steps S140 and S150) of calculating an error between the frequency distribution of the original data and the frequency distribution of the extracted data, and determining whether the original data and the extracted data are similar to each other. The processing circuit 510 executes a search process in which the processes from step 1 to step 4 are repeatedly executed while changing the settings of a plurality of time windows. Then, the processing circuit 510 stores in the storage device 520 a data extraction pattern when extracted data similar to the original data is extracted (step S160).

[0071] The data center 500 uses the rotational speed of the second oil pump 57, which is an oil pump that discharges lubricating oil to lubricate the components of the power split mechanism 40, as a feature. The oil pump discharges more lubricating oil as the rotational speed of the oil pump increases. The data center 500 determines the similarity between the original data and the extracted data using the frequency distribution of the feature. The data center 500 extracts extracted data in which the frequency distribution of the feature in the original data is similar to the frequency distribution of the feature. In other words, the data center 500 extracts extracted data in which the distribution of the rotational speed of the second oil pump 57 is similar to the original data. Therefore, when analyzing the lubrication history using the extracted data, analysis results similar to those obtained using the original data can be obtained. The data center 500 can extract data from the original data that is suitable for analyzing the lubrication history of the components of the power split mechanism 40 installed in the vehicle 10.

[0072] <Effects of the second embodiment> The second embodiment has the following advantages similar to the advantage (1-1) of the first embodiment. (2-1) According to the data center 500, data suitable for analyzing the lubrication history of components in the power transmission device mounted on the vehicle 10 can be extracted from the original data.

[0073] <Modification of the second embodiment> As in the first embodiment, the process of step S125 for dividing the feature amounts by the second feature amount may be executed, and an extraction pattern may be searched for using the first feature amount and the second feature amount.

[0074] <Example of change> Common modifiable elements of the above embodiments include the following: The following modifications can be implemented in combination with each other within the scope of technical compatibility.

[0075] In the above embodiment, an example has been given in which the information processing device is embodied as the data center 500. However, the information processing device may be embodied as the information processing terminal 600. Furthermore, the information processing device may be embodied as a control device for the vehicle 10.

[0076] The data center 500 determines the similarity between the original data and the extracted data by calculating the error in the frequency distribution. Alternatively, the data center 500 may determine whether the original data and the extracted data are similar without calculating the error. For example, by using a statistical method such as a goodness-of-fit test, it can be determined that the original data and the extracted data are similar when the difference between them is not significant. [Explanation of symbols]

[0077] 10...vehicle, 20...hybrid mechanism, 30...engine, 31...first motor generator, 32...second motor generator, 40...power split mechanism, 41...case, 42...planetary gear unit, 43...planetary carrier, 44...ring, 45...sun gear, 46...pinion gear, 47...ring gear, 49...reduction gear, 53...differential ring gear, 56...first oil pump, 57...second oil pump, 80...vehicle control unit, 90...communication device, 400...communication network, 500...data center, 510...processing circuit, 520...storage device, 530...communication device, 600...information processing terminal, 610...processing circuit, 620...storage device, 630...communication device

Claims

1. An information processing device that extracts a portion of data from original data collected over a predetermined period of time using a plurality of sensors mounted on a vehicle, and extracts data for analyzing the lubrication history of components in a power transmission device mounted on the vehicle, the information processing device comprising: a processing circuit; The processing circuit repeatedly executes the following steps by changing the settings of the multiple time windows: a first step of setting multiple time windows to extract data for a portion of the original data so that the sum of the periods of all the time windows is shorter than the predetermined period; a second step of extracting data from the original data using the multiple time windows; a third step of calculating a frequency distribution of the feature obtained from the original data and a frequency distribution of the feature obtained from extracted data obtained by combining all of the data extracted using the multiple time windows, using the rotational speed of a rotating body that lubricates the components by rotating and scooping up lubricating oil stored in a case of the power transmission device; and a fourth step of determining whether the original data and the extracted data are similar to each other using the frequency distribution, while changing the settings of the multiple time windows, and extracts the extracted data that is similar to the original data. Information processing device.

2. An information processing device that extracts a portion of data from original data collected over a predetermined period of time using a plurality of sensors mounted on a vehicle, and extracts data for analyzing the lubrication history of components in a power transmission device mounted on the vehicle, the information processing device comprising: a processing circuit; The processing circuit repeatedly executes the following steps by changing the settings of the multiple time windows: a first step of setting multiple time windows to extract data for a portion of the original data so that the sum of the periods of all the time windows is shorter than the predetermined period; a second step of extracting data from the original data using the multiple time windows; a third step of calculating a frequency distribution of a feature obtained from the original data, using the rotational speed of an oil pump that discharges lubricating oil to lubricate the component, and a frequency distribution of the feature obtained from extracted data that combines all of the data extracted using the multiple time windows; and a fourth step of determining whether the original data and the extracted data are similar to each other by changing the settings of the multiple time windows, and extracts the extracted data that is similar to the original data. Information processing device.

3. the processing circuit classifies the original data into a plurality of data by the second feature amount, the second feature amount being a physical quantity related to the lubrication history of the component part, the second feature amount being different from the feature amount included in the original data, and calculates a frequency distribution of the feature amount obtained from the original data for each of the classified data, and also classifies the extracted data into a plurality of data by the second feature amount in correspondence with the classification of the plurality of data in the original data, and calculates a frequency distribution of the feature amount obtained from the extracted data for each of the classified data, For each of the data segments obtained in the fourth step, an error between the frequency distribution of the original data and the frequency distribution of the extracted data is calculated, and when all of the errors for each segment are equal to or smaller than a threshold, the original data and the extracted data are determined to be similar.

3. The information processing device according to claim 1 or 2.

4. The processing circuit defines the rotation speed of a second rotating body that rotates in a case of the power transmission device and scatters adhering lubricating oil to lubricate the component as the second characteristic amount. The information processing device according to claim 3 .

5. The processing circuitry The tilt angle of the vehicle is set as the second feature amount. The information processing device according to claim 3 .

Images