Failure prediction system

Abstract

To provide a failure prediction system capable of improving determination accuracy of failure prediction and reducing a data communication amount between a vehicle and a server.SOLUTION: A failure prediction system, which determines failure prediction with respect to an observation object including the failure prediction system and a component configuring the same, comprises: an acquisition section which acquires time series data representing an operation state of the observation object; a determination section which determines whether the operation state of the observation object meets a predetermined condition; an identification section which identifies a statistic amount of the time series data during a period when the operation state meets the predetermined condition; and a symptom determination section which determines failure prediction of the observation object based on the identified statistic amount.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present disclosure relates to a failure prediction system. [Background technology]

[0002] Conventionally, a failure prediction system is known that obtains actual measured values ​​from the vehicle that measure the operating state of an object to be observed, including a system installed in the vehicle and components that constitute the system, and determines whether or not a component will fail based on the deviation between the actual measured value and a target value using a threshold value.

[0003] For example, Patent Document 1 describes a failure prediction device for a sensor provided in an exhaust system of an engine mounted on a vehicle, the failure prediction device including a mileage calculation unit that calculates a cumulative mileage of the vehicle, an idling operation time calculation unit that calculates a cumulative idling operation time of the engine, and a sensor failure occurrence prediction unit that predicts the occurrence of a sensor failure based on the calculated cumulative mileage and the calculated cumulative idling operation time. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2018-148720 A Summary of the Invention [Problem to be solved by the invention]

[0005] However, when the operating state of the observed object does not satisfy a predetermined condition, the accuracy of the failure prediction judgment of the observed object may be reduced if the failure prediction judgment is performed based on the actual measurement value.

[0006] In addition, in cases where actual measured values ​​are sent from the vehicle to a server and the server determines a failure prediction for the observed object based on the actual measured values, and a limit is placed on the amount of data communication between the vehicle and the server, if the amount of data communication exceeds the limit, the number of vehicles for which a failure prediction determination is performed may be limited.

[0007] An object of the present disclosure is to provide a failure prediction system that can improve the accuracy of failure prediction judgments and reduce the amount of data communication between vehicles and servers. [Means for solving the problem]

[0008] In order to achieve the above object, the failure prediction system in the present disclosure includes: A failure prediction system for determining a failure prediction of an observation object including a system and a component constituting the system, An acquisition unit that acquires time-series data indicating a motion state of the observation target; A determination unit that determines whether or not the motion state of the observation target satisfies a predetermined condition; an identification unit that identifies a statistic of the time-series data during a period in which the operational state satisfies the predetermined condition; a sign determination unit that determines a failure prediction of the observation target based on the specified statistics; Equipped with. Effect of the Invention

[0009] According to the present disclosure, it is possible to improve the accuracy of failure prediction and reduce the amount of data communication between the vehicle and the server. [Brief description of the drawings]

[0010] [Figure 1] FIG. 1 is a diagram illustrating an example of a configuration of a control device and the like according to an embodiment of the present disclosure. [Diagram 2] FIG. 2 is a diagram showing an overall configuration of a failure prediction system according to an embodiment of the present disclosure. [Diagram 3]3 shows the difference between the actual measured value and the target value in the upper part, and the permission flag in the lower part. [Figure 4] FIG. 4 is a diagram showing statistics of actual measured values ​​for each day. [Diagram 5] FIG. 5 is a diagram showing an example of the malfunction level value. [Figure 6] FIG. 6 is a flowchart illustrating an example of the failure prediction determination process. [Figure 7] FIG. 7 is a diagram showing the difference between the characteristic value and the diagnostic threshold value. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. FIG. 1 is a block diagram showing a schematic configuration of a control device and the like in an embodiment of the present disclosure. In FIG. 1, arrows indicate main data flows, and there may be data flows not shown in FIG. 1. In FIG. 1, each functional block shows a configuration in function units, not a configuration in hardware (device) units. Therefore, the functional blocks shown in FIG. 1 may be implemented in a single device, or may be implemented separately in multiple devices. Data may be exchanged between functional blocks via any means, such as a data bus or a communication network.

[0012] FIG. 2 is a diagram showing the overall configuration of a failure prediction system in an embodiment of the present disclosure. The failure prediction system in the embodiment of the present disclosure judges failure prediction of a system and components constituting the system. In the following description, the system and components constituting the system are referred to as "observation target." Note that the observation target in this embodiment is mounted on a vehicle. The observation target includes, for example, an engine system, a transmission system, and an aftertreatment system, and the components constituting each system. The operating state of the observation target is measured by a predetermined sensor or the like (not shown). In the following description, the measurement value measured by the sensor is referred to as an "actual measurement value."

[0013] 2 shows an engine controller 1A which is a controller for the engine system, a transmission controller 1B which is a controller for the transmission system, and an aftertreatment controller 1C which is a controller for the aftertreatment system. Each of the controllers 1A, 1B, and 1C judges whether or not the operating state of the observed object satisfies a predetermined condition, and transmits the judgment result to the control device 2.

[0014] 1, a failure prediction system 100 in this embodiment includes a control device 2, a communication device 3, and a server 4. The control device 2 and the communication device 3 are mounted on a vehicle.

[0015] The server 4 has a communication unit 5 and a sign determination unit 6. The communication unit 5 transmits and receives data to and from the communication device 3. The sign determination unit 6 determines failure prediction of the observed object based on data transmitted from the vehicle side (sign determination). In the following description, the data transmitted from the vehicle side may be referred to as "statistics." The failure prediction system 100 in this embodiment transmits data as statistics from the vehicle side to the server 4, thereby making it possible to communicate data over a wide range and reduce the communication load. The "statistics" will be described in detail later.

[0016] The control device 2 has a control unit 30 and a storage unit 36. When sign determination is possible, sign determination permission information is transmitted from each of the controllers 1A, 1B, 1C to the control device 2 (CAN transmission). In addition, actual measurement values ​​obtained by measuring the operating state of the observation target are transmitted from each of the controllers 1A, 1B, 1C to the control device 2 (CAN transmission). In the following description, the actual measurement values ​​may be referred to as "time series data" or "frequency data".

[0017] (Storage unit 36) The memory unit 36 ​​is a storage device such as a ROM (Read Only Memory) that stores the BIOS (Basic Input Output System) of the computer that realizes the control device 2, a RAM (Random Access Memory) that serves as the working area of ​​the control device 2, an HDD (Hard Disk Drive) or SSD (Solid State Drive) that stores the OS (Operating System), application programs, and various information referenced when the application programs are executed.

[0018] The control unit 30 is a processor such as a CPU (Central Processing Unit) or GPU (Graphics Processing Unit) of the control device 2, and functions as an acquisition unit 31, a judgment unit 32, a difference calculation unit 33, an accumulation unit 34, and an identification unit 35 by executing programs stored in the memory unit 36.

[0019] 1 shows an example in which the control device 2 is configured as a single device. However, the control device 2 may be realized by, for example, multiple processors, memories, and other computational resources. In this case, each part constituting the control unit 30 is realized by at least one of multiple different processors executing a program.

[0020] As described above, each of the controllers 1A, 1B, and 1C determines whether or not a specified condition is met, and if the specified condition is met, transmits on information of the permission flag to the control device 2, and if the specified condition is not met, transmits off information of the permission flag to the control device 2.

[0021] (Acquisition part 31) The acquisition unit 31 acquires time-series data (frequency data) for a period during which a predetermined condition is satisfied (a period during which the permission flag is in an on state). Note that satisfaction of the predetermined condition means that the operating states of the engine, transmission, and aftertreatment, which are the objects of observation, satisfy thresholds in accordance with a predetermined fault diagnosis logic.

[0022] (Specific examples of specified conditions) Next, a specific example of the predetermined condition will be described. First, the predetermined conditions in the engine will be described. The engine controller 1A monitors the following parameters, and when each parameter is within a predetermined range or in a predetermined state, it determines that a predetermined condition is satisfied. -Engine fuel flow rate is within a specified range -Engine speed is within a specified range -Engine exhaust temperature is within a specified range - The magnitude of engine vibration is within a specified range. -Engine torque fluctuation is within a specified range -Engine oil consumption is within the specified range

[0023] Next, the predetermined conditions in the transmission will be described. The transmission controller 1B monitors the following parameters, and when each parameter is within a predetermined range or in a predetermined state, it determines that a predetermined condition is satisfied. - The difference between the inner input shaft rotation speed and the turbine rotation speed is within a specified range. - Ignition voltage is within the specified range The current gear is the specified gear. -The target gear and current gear must match - The specified number of seconds or more have passed since the odd-numbered gear clutch was engaged. - The inner input shaft rotation speed is not 0 rpm. - Turbine speed is not 0 rpm

[0024] Next, the predetermined conditions in the post-processing will be described. The aftertreatment controller 1C monitors the following parameters, and when each parameter is within a predetermined range or in a predetermined state, it determines that a predetermined condition is satisfied. - NOx conversion efficiency is within a specified range - Battery voltage is within the specified range The status switch is on. The engine is running. - Intake air temperature is within a specified range Atmospheric pressure is within a specified range DPF (Diesel Particulate Filter) regeneration is stopped. · SCR (Selective Catalytic Reduction) catalyst temperature is within a specified range - Exhaust gas flow rate is within a specified range · NOx sensor 1 upstream of the SCR catalyst and NOx sensor 2 downstream of the SCR catalyst are valid. The concentration of NOx sensor 1 is within the specified range. The DEF supply module is on.

[0025] (Judgment unit 32) The determination unit 32 determines whether or not the permission flag is in an on state (a state in which the operating state of the observation target satisfies a threshold value according to a predetermined fault diagnosis logic). When the permission flag is in an on state, the control unit 30 stores the acquired time-series data in the storage unit 36.

[0026] (Difference calculation unit 33) The difference calculation unit 33 calculates the difference between the actual measurement value and a predetermined target value. In the following description, the difference calculated by the difference calculation unit 33 may be referred to as a "characteristic value." The characteristic value may also be referred to as "time series data" or "frequency data."

[0027] In Fig. 3, the upper part shows the difference (characteristic value) between the actual measurement value and the target value, and the lower part shows the permission flag. In the upper part of Fig. 3, the actual measurement value is shown by a thick solid line, the target value is shown by a thin solid line, and the difference (characteristic value) is indicated by A. In the lower part of Fig. 3, the ON state of the permission flag is indicated by hatched time periods T1 and T2. That is, the difference calculation unit 33 calculates the difference based on the actual measurement values ​​and the target values ​​acquired in each of the time periods T1 and T2 in which the permission flag is in the ON state.

[0028] (Specific Section 35) The determination unit 35 calculates statistics of the time series data. The control unit 30 stores the statistics in the storage unit 36. Here, the time series data is data for one day. The determination unit 35 calculates statistics for one day based on the time series data (actual measurements and characteristic values) for one day. The control unit 30 stores the calculated daily statistics in the storage unit 36. The determination unit 35 calculates the daily statistics periodically, for example, daily, weekly, or monthly. The statistics stored in the storage unit 36 ​​are transmitted from the communication device 3 to the server 4 (communication unit 5). The communication device 3 transmits the daily statistics periodically, for example, daily, weekly, or monthly.

[0029] (statistics) Next, the daily statistics transmitted from the vehicle side to the server 4 will be described with reference to FIG. 4. FIG. 4 is a diagram showing the daily statistics of the actual measurement values. The daily statistics of the actual measurement values ​​include a standard value, an average value, a standard deviation, a minimum value, a maximum value, and an outlier count. Here, the "standard value" is a numerical value at the center position in a normal distribution. The "average value" is a numerical value obtained by dividing the total of the actual measurement values ​​for one day by the number of the actual measurement values. The "diagnosis threshold" is a threshold at which a fault is diagnosed, and is also called a fault threshold (see FIG. 5). The "premonition threshold" is a threshold that indicates a premonition of a fault, and is also called a premonition determination value (see FIG. 5). The "minimum value" is the minimum value among the actual measurement values ​​for one day. The "maximum value" is the maximum value among the actual measurement values ​​for one day. The "outlier count" is the number of times that the actual measurement values ​​for one day deviate from the premonition threshold, if any.

[0030] The daily statistics transmitted from the vehicle to the server 4 are numbered No.1, No.2, No.3, No.4, etc. In FIG. 4, the average value "***", standard deviation "***", minimum value "***", maximum value "***", and outlier count "***" are shown as the daily statistics corresponding to No.1. The average value "***", standard deviation "***", minimum value "***", maximum value "***", and outlier count "***" are shown as the daily statistics for No.2. The average value "***", standard deviation "***", minimum value "***", maximum value "***", and outlier count "***" are shown as the daily statistics for No.3. The average value "***", standard deviation "***", minimum value "***", maximum value "***", and outlier count "***" are shown as the daily statistics for No.4. The "***" indicates a specific value that has been masked.

[0031] (Other statistics) As described above, the daily statistics transmitted from the vehicle to the server 4 are statistics of actual measurements, but the present disclosure is not limited to this, and may be, for example, the difference (characteristic value) shown in FIG.

[0032] FIG. 5 is a diagram showing an example of ranking values. The horizontal axis of FIG. 5 indicates time, and the vertical axis indicates malfunction level values. FIG. 5 also indicates the precursor judgment value with a dashed line, and the failure threshold with a dashed line. Also, FIG. 5 shows a malfunction level value Ra that exceeds the precursor judgment value, and a malfunction level value Rb that falls below the precursor judgment value. The "malfunction level value" is a numerical value that indexes the margin of a statistic with respect to the failure threshold (also called a diagnosis threshold, see FIG. 4).

[0033] (Omen Judgment Unit 6) As shown in FIG. 5, the sign determining unit 6 calculates a malfunction level value every month based on the statistics transmitted from the vehicle side to the server 4.

[0034] The sign determination unit 6 determines whether the calculated malfunction level value is below the sign determination value. When the malfunction level value exceeds the sign determination value, such as the malfunction level value Ra shown in Fig. 5, the sign determination unit 6 determines that there is no sign of failure in the observed object. On the other hand, when the malfunction level value is below the sign determination value, such as the malfunction level value Rb shown in Fig. 5, the sign determination unit 6 determines that there is a sign of failure in the observed object.

[0035] (Failure prediction and judgment processing) Next, an example of the failure prediction judgment process will be described with reference to Fig. 6. Fig. 6 is a flowchart showing an example of the failure prediction judgment process. Here, the time series data of the observation target transmitted from each of the controllers 1A, 1B, and 1C of each system will be described as actual measurement values.

[0036] Each of the controllers 1A, 1B, 1C determines whether the motion state of the observation target satisfies a predetermined condition, and if the predetermined condition is satisfied, transmits sign determination permission information (permission flag ON) to the control device 2 (step S1).

[0037] The determination unit 32 determines whether or not the sign determination permission information has been transmitted, and if the sign determination permission information has been transmitted, the acquisition unit 31 acquires the actual measurement values ​​of the observation target transmitted from each of the controllers 1A, 1B, and 1C. Furthermore, the difference calculation unit 33 calculates the difference between the actual measurement value and the target value (step S2).

[0038] The identification unit 35 calculates statistics of the actual measured values ​​for one day (standard value, average value, standard deviation value, minimum value, maximum value, outlier count). The control unit 30 stores the statistics of the actual measured values ​​for one day in the storage unit 36. The communication device 3 also transmits the statistics of the actual measured values ​​for one day to the communication unit 5 of the server 4 (step S3).

[0039] The communication unit 5 receives the statistics. The sign determination unit 6 calculates a malfunction level value based on the statistics, and performs a failure prediction determination for the observation target based on the calculated malfunction level value (sign determination) (step S4).

[0040] The failure prediction system in the above embodiment is a failure prediction system that performs a failure prediction judgment on an object of observation including the system and components that constitute the system, and includes an acquisition unit 31 that acquires time series data indicating the operating state of the object of observation, a judgment unit 32 that judges whether the operating state of the object of observation satisfies a predetermined condition, an identification unit 35 that identifies statistics of the time series data for a period during which the operating state satisfies the predetermined condition, and a sign judgment unit 6 that performs a failure prediction judgment on the object of observation based on the identified statistics.

[0041] With the above configuration, since failure prediction is determined based on time-series data for a period that satisfies a predetermined condition, it is possible to improve the accuracy of failure prediction. Also, since the sign determination unit 6 determines failure prediction of the observed object based on statistics, it is possible to reduce the amount of data communication by transmitting the statistics of the time-series data to the sign determination unit 6 compared to transmitting the time-series data to the sign determination unit 6.

[0042] In the failure prediction system according to the above embodiment, the objects to be observed include the engine system, the transmission system, and the aftertreatment system. This makes it possible to improve the accuracy of failure prediction for the engine system and the like, and also to reduce the amount of data communication required for failure prediction for the engine system and the like.

[0043] In addition, in the failure prediction system in the above embodiment, the objects to be observed include the engine, the transmission, and the aftertreatment device. This makes it possible to improve the accuracy of failure prediction of the engine, etc., and also to reduce the amount of data communication for failure prediction of the engine, etc.

[0044] In the failure prediction system according to the above embodiment, the predetermined condition is a condition that satisfies a threshold value according to a predetermined failure diagnosis logic. This makes it possible to accurately grasp the signs of failure, thereby improving the accuracy of failure prediction.

[0045] Moreover, in the failure prediction system in the above embodiment, the statistical quantity is at least one of the average value, standard deviation, minimum value, and maximum value of the time series data. This significantly reduces the amount of data transmitted from the vehicle to the server 4, making it possible to significantly reduce the amount of data communication transmitted to the symptom determination unit 6.

[0046] In addition, in the failure prediction system according to the above embodiment, the time-series data includes a characteristic value indicating the difference between an actual measurement value of the operating state of the observed object and a predetermined target value, which makes it possible to more accurately grasp a sign of failure based on the difference.

[0047] Furthermore, the failure prediction system in the above embodiment includes a vehicle having an acquisition unit 31, a determination unit 32, and an identification unit 35, and a server 4 having a symptom determination unit 6, and statistics are transmitted from the vehicle to the server 4. This makes it possible to reduce the amount of data communication between the vehicle and the server.

[0048] (Modification) In the above embodiment, the actual measurement value and the characteristic value (target value-actual measurement value) are shown as the statistical quantity, but the present disclosure is not limited thereto. For example, the statistical quantity may include an integrated value obtained by integrating the difference between the diagnostic threshold value and the characteristic value. Here, the "diagnostic threshold value" is a threshold value that is changed according to the operating status and environment of the system. The integration of the difference between the diagnostic threshold value and the characteristic value is performed by the integration unit 34 (see FIG. 1).

[0049] FIG. 7 shows the difference between the characteristic value and the diagnostic threshold in the upper part, and the permission flag in the lower part. In the upper part of FIG. 7, the characteristic value is shown by a thick solid line, the diagnostic threshold is shown by a dashed line, and the integrated value of the difference is shown by B. In the lower part of FIG. 7, the ON state of the permission flag is shown by the shaded time periods T1 and T2. That is, the integration unit 34 integrates the difference in each of the time periods T1 and T2 in which the permission flag is ON. The integrated value for each day obtained by integrating the difference is transmitted from the vehicle side to the server 4. The sign determination unit 6 calculates a malfunction level value for each month based on the daily integrated value, and determines whether or not the observed object has a malfunction sign based on the malfunction level value. The integrated value is the integrated value of the difference between the characteristic value and a predetermined diagnostic threshold, and the diagnostic threshold is a threshold that is changed according to the operating conditions and environment of the system, so that it is possible to accurately determine the malfunction prediction based on the integrated value without being affected by the operating conditions and environment of the system.

[0050] In addition, the above-mentioned embodiments are merely examples of the embodiment of the present disclosure, and the technical scope of the present disclosure should not be interpreted as being limited by them. In other words, the present disclosure can be embodied in various forms without departing from the gist or main features thereof. [Industrial Applicability]

[0051] The present disclosure is preferably used in a vehicle management system equipped with a failure prediction system that is required to improve the accuracy of failure prediction judgment and reduce the amount of data communication between the vehicle and the server. [Explanation of symbols]

[0052] 1A Engine Controller 1B Transmission Controller 1C Aftertreatment Controller 2. Control device 3. Communications equipment 4 Server 5. Communications Department 6 Prediction Unit 30 Control section 31 Acquisition Department 32 Judgment section 33 Difference calculation part 34 Integration section 35 Specific part 36 Memory section 100 Failure Prediction System

Claims

1. A failure prediction system for determining a failure prediction of an observation object including a system and a component constituting the system, An acquisition unit that acquires time-series data indicating a motion state of the observation target; A determination unit that determines whether or not the motion state of the observation target satisfies a predetermined condition; an identification unit that identifies a statistic of the time-series data during a period in which the operational state satisfies the predetermined condition; a sign determination unit that determines a failure prediction of the observation target based on the specified statistics; Equipped with Failure prediction system.

2. The observed object includes an engine system, a transmission system, and an aftertreatment system. The failure prediction system according to claim 1 .

3. The object to be observed includes an engine, a transmission, and an aftertreatment device. The failure prediction system according to claim 1 .

4. The predetermined condition is a condition that satisfies a threshold value according to a predetermined fault diagnosis logic. The failure prediction system according to any one of claims 1 to 3.

5. The statistical quantity is at least one of an average value, a standard deviation, a minimum value, and a maximum value of the time series data. The failure prediction system according to any one of claims 1 to 4.

6. The time-series data includes a characteristic value indicating a difference between an actual measurement value obtained by measuring the operating state of the observation target and a predetermined target value. The failure prediction system according to any one of claims 1 to 5.

7. The time series data includes an integrated value obtained by integrating a difference between the characteristic value and a predetermined diagnostic threshold value. The failure prediction system according to claim 6.

8. The diagnostic threshold is a predetermined value depending on the driving situation and environment of the object to be observed. The failure prediction system according to claim 7.

9. a vehicle having the acquisition unit, the determination unit, and the identification unit; A server having the sign determination unit; Equipped with The statistics are transmitted from the vehicle to the server. The failure prediction system according to claim 1 .

Images