Information processing device

Abstract

To provide an information processing device that facilitates discovering a diagnostic pattern suitable for abnormality diagnosis.SOLUTION: An information processing device 100 determines a diagnostic pattern to be used in an abnormality diagnosis system. The information processing device 100 repeats a simulation of processing for diagnosing an abnormality of a hydraulic control device by using a simulation device 200 storing a model M_240 for the hydraulic control device that can simulate the operation of the hydraulic control device, thereby searching for a diagnostic pattern suitable for the processing for diagnosing an abnormality.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an information processing device that determines a diagnostic pattern to be used in an abnormality diagnostic system for a hydraulic control device. [Background technology]

[0002] Patent Document 1 discloses that the degree of anomaly is calculated based on time-series data acquired by a plurality of sensors, and the state of the target data is determined. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2019-096014 Summary of the Invention [Problem to be solved by the invention]

[0004] When diagnosing an abnormality in a hydraulic control device based on time-series data obtained by controlling the hydraulic control device using a predetermined diagnostic pattern, it is desirable to operate the hydraulic control device using a diagnostic pattern suitable for diagnosing the abnormality and obtain the data to be used for diagnosis. [Means for solving the problem]

[0005] The means for solving the above problems and their effects will be described below. An information processing device for solving the above problem determines a diagnostic pattern to be used in an abnormality diagnosis system. The abnormality diagnosis system includes a processing circuit and a storage device. The storage device of the abnormality diagnosis system stores reference cluster data obtained by classifying differential pressure data at multiple times, included in data on the transition of the differential pressure between the hydraulic pressure and the target hydraulic pressure when a hydraulic control device not experiencing an abnormality is operated according to a predetermined diagnostic pattern, into a predetermined number of clusters using clustering, a type of machine learning. The processing circuit of the abnormality diagnosis system creates evaluation data, which is data indicating the transition of the differential pressure between the hydraulic pressure and the target hydraulic pressure, recorded while the hydraulic control device to be evaluated is operated according to the diagnostic pattern. The processing circuit of the abnormality diagnosis system compares the differential pressure data at multiple times included in the evaluation data with the centers of gravity of each cluster in the reference cluster data, and creates evaluation cluster data by determining to which of the predetermined number of clusters in the reference cluster data each piece of data at multiple times in the evaluation data belongs. The processing circuit of the abnormality diagnosis system determines that an abnormality has occurred in the hydraulic control device based on a discrepancy between the cluster determination results in the reference cluster data and the evaluation cluster data. The information processing device uses a simulation device that stores a model of the hydraulic control device capable of simulating the operation of the hydraulic control device to repeatedly simulate a process for diagnosing an abnormality in the hydraulic control device. By repeating the simulation, the information processing device searches for the diagnostic pattern that is suitable for the process for diagnosing the abnormality. [Effects of the Invention]

[0006] The information processing device can easily perform a search under various settings rather than searching for a diagnostic pattern by repeatedly operating an actual hydraulic control device, and therefore the information processing device can easily find a diagnostic pattern suitable for diagnosing an abnormality. [Brief explanation of the drawings]

[0007] [Figure 1]FIG. 1 is a schematic diagram showing an embodiment of an information processing device and a simulation device. [Figure 2] FIG. 2 is a schematic diagram showing the configuration of the hydraulic control device. [Figure 3] FIG. 3 is a graph showing an example of a diagnostic pattern. [Figure 4] FIG. 4 is a graph showing an example of clustering. [Figure 5] FIG. 5 is a flowchart showing the flow of processing executed by the abnormality diagnosis system. [Figure 6] FIG. 6 is a table showing the clustering results. [Figure 7] FIG. 7 is a table showing the relationship between the clustering results and the type of anomaly. [Figure 8] FIG. 8 is a flowchart showing the flow of processing executed by the information processing device. DETAILED DESCRIPTION OF THE INVENTION

[0008] An embodiment of an information processing device will be described below with reference to FIGS. <Configuration of information processing device 100> 1 shows an information processing device 100 and a simulation device 200. The simulation device 200 stores a model M_240 that can simulate the operation of a hydraulic control device. Hereinafter, the hydraulic control device simulated by the model M_240 will be referred to as the hydraulic control device 240.

[0009] The information processing device 100 searches for and determines a diagnostic pattern to be used by the abnormality diagnosis system using the simulation device 200. The abnormality diagnosis system diagnoses the hydraulic control device 240 by operating the hydraulic control device 240 using the predetermined diagnostic pattern.

[0010] 1, information processing device 100 includes processing circuitry 110, storage device 120, and display 130. Processing circuitry 110 includes a CPU that executes processing according to a program and a ROM in which the program is stored. Storage device 120 stores data. Information processing device 100 is, for example, a personal computer or a workstation.

[0011] <Configuration of hydraulic control device 240> Next, the configuration of the hydraulic control device 240 simulated by the model M_240 will be described with reference to FIG.

[0012] The hydraulic control device 240 is, for example, part of a continuously variable transmission mounted on an automobile. The continuously variable transmission transmits power by a belt wound around a first pulley 210 and a second pulley 220. The hydraulic control device 240 controls a first hydraulic pressure Pp supplied to the first pulley 210 and a second hydraulic pressure Ps supplied to the second pulley 220 of the continuously variable transmission. The hydraulic control device 240 changes the gear ratio of the continuously variable transmission by controlling the first hydraulic pressure Pp and the second hydraulic pressure Ps.

[0013] 2, the hydraulic control device 240 includes a regulator valve 241, a modulator valve 242, a first control valve 247, and a second control valve 248. The hydraulic control device 240 also includes a first linear solenoid valve 243, a second linear solenoid valve 244, a first hydraulic damper 245, and a second hydraulic damper 246.

[0014] The regulator valve 241 adjusts the hydraulic pressure of the oil supplied from the oil pump. The hydraulic pressure adjusted by the regulator valve 241 is the line pressure PL. The oil adjusted to the line pressure PL is supplied to a first control valve 247 and a second control valve 248. The hydraulic pressure adjusted to the line pressure PL is also supplied to a modulator valve 242.

[0015] The first control valve 247 supplies oil adjusted to the line pressure PL to the first pulley 210. The first control valve 247 controls the amount of oil supplied to the first pulley 210. The first control valve 247 increases the first oil pressure Pp by supplying oil adjusted to the line pressure PL to the first pulley 210. The first control valve 247 controls the amount of oil discharged from the first pulley 210. The first control valve 247 decreases the first oil pressure Pp by discharging oil from the first pulley 210.

[0016] The second control valve 248 supplies oil adjusted to the line pressure PL to the second pulley 220. The second control valve 248 controls the amount of oil supplied to the second pulley 220. The second control valve 248 increases the second oil pressure Ps by supplying oil adjusted to the line pressure PL to the second pulley 220. The second control valve 248 also controls the amount of oil discharged from the second pulley 220. The second control valve 248 decreases the second oil pressure Ps by discharging oil from the second pulley 220.

[0017] Increasing the first hydraulic pressure Pp and decreasing the second hydraulic pressure Ps increases the winding radius of the belt around the first pulley 210 and decreases the winding radius of the belt around the second pulley 220. Increasing the second hydraulic pressure Ps and decreasing the first hydraulic pressure Pp increases the winding radius of the belt around the second pulley 220 and decreases the winding radius of the belt around the first pulley 210. In this way, the hydraulic control device 240 changes the gear ratio of the continuously variable transmission.

[0018] The first control valve 247 and the second control valve 248 are each controlled by a signal oil pressure. The modulator valve 242 adjusts the line pressure PL to the signal oil pressure. The oil adjusted to the signal oil pressure is supplied to the first linear solenoid valve 243 and the second linear solenoid valve 244.

[0019] The first linear solenoid valve 243 controls the signal oil pressure supplied to the first control valve 247. The second linear solenoid valve 244 controls the signal oil pressure supplied to the second control valve 248. The signal oil pressure supplied to the first control valve 247 is controlled by opening and closing the first linear solenoid valve 243. The signal oil pressure supplied to the second control valve 248 is controlled by opening and closing the second linear solenoid valve 244.

[0020] A first hydraulic damper 245 is provided in the oil passage connecting the first linear solenoid valve 243 and the first control valve 247. A second hydraulic damper 246 is provided in the oil passage connecting the second linear solenoid valve 244 and the second control valve 248.

[0021] <Abnormality detection process> When diagnosing the hydraulic control device 240, a plurality of hydraulic sensors that detect the hydraulic pressure of each part of the hydraulic control device 240 are connected to the hydraulic control device 240. The diagnosis of the hydraulic control device 240 is performed, for example, as a pre-shipment inspection after manufacture.

[0022] An abnormality diagnosis system that diagnoses the hydraulic control device 240 creates evaluation data that indicates the transition of the differential pressure between the hydraulic pressure and the target hydraulic pressure, which is recorded while the hydraulic control device 240 to be evaluated is operated in a predetermined diagnostic pattern. The abnormality diagnosis system creates evaluation cluster data by determining to which cluster in the reference cluster data each piece of data in the evaluation data belongs. The processing circuit 110 executes abnormality determination processing that determines that an abnormality has occurred in the hydraulic control device 240 based on the fact that the cluster determination results are different.

[0023] The predetermined diagnostic pattern divides time periods into periods according to the type of abnormality to be detected, and sequentially controls the hydraulic pressures controlled by the hydraulic control device 240 in a pattern that makes it easy to detect each abnormality. The abnormality diagnosis system detects the type of abnormality occurring in the hydraulic control device 240 based on information about the period of time in which the reference cluster data and the evaluation cluster data diverge.

[0024] The predetermined diagnostic pattern is defined by a target hydraulic pressure PLt for the line pressure PL, a target hydraulic pressure Ppt for the first hydraulic pressure Pp, and a target hydraulic pressure Pst for the second hydraulic pressure Ps. The predetermined diagnostic pattern is defined so that abnormalities in each part of the hydraulic control device 240 can be diagnosed from the transition of the measured hydraulic pressures by varying the line pressure PL, the first hydraulic pressure Pp, and the second hydraulic pressure Ps over a predetermined period.

[0025] The abnormality diagnosis system controls multiple hydraulic pressures according to a diagnosis pattern so that each abnormality can be easily diagnosed. Therefore, if the time period during which the reference cluster data and the evaluation cluster data diverge is known, it can be determined that an abnormality corresponding to that time period has occurred. The abnormality diagnosis system determines the type of abnormality occurring in the hydraulic control device 240 based on information indicating the time period during which the reference cluster data and the evaluation cluster data diverge.

[0026] <Reference cluster data> Reference cluster data is stored in the storage device of the abnormality diagnosis system. The reference cluster data is data resulting from classifying, into a predetermined number of clusters, differential pressure data at multiple times included in data on the transition of the differential pressure between the hydraulic pressure and the target hydraulic pressure when a hydraulic control device 240 having no abnormality is operated in a predetermined diagnostic pattern. To create the reference cluster data, the hydraulic pressure is measured while a hydraulic control device 240 having no abnormality is operated in a predetermined diagnostic pattern. For example, of the hydraulic pressures at various locations in the hydraulic control device 240, the line pressure PL, the first hydraulic pressure Pp, and the second hydraulic pressure Ps are measured.

[0027] 3 shows the transition of the target hydraulic pressure Ppt of the first hydraulic pressure Pp in a predetermined diagnostic pattern. In the predetermined diagnostic pattern, the target hydraulic pressure PLt of the line pressure PL, the target hydraulic pressure Ppt of the first hydraulic pressure Pp, and the target hydraulic pressure Pst of the second hydraulic pressure Ps all have different fluctuation patterns.

[0028] The actual hydraulic pressure fluctuates with a delay relative to fluctuations in the target hydraulic pressure. The actual hydraulic pressure may overshoot the target hydraulic pressure. The actual hydraulic pressure may not immediately converge to the target hydraulic pressure, but may fluctuate above and below the target hydraulic pressure until it converges to the target hydraulic pressure. Data showing the trend in the differential pressure between the target hydraulic pressure and the measured hydraulic pressure includes information on the delay, overshoot, and oscillation of the actual hydraulic pressure relative to the target hydraulic pressure. Such delay, overshoot, and oscillation also occur in a hydraulic control device 240 that is not experiencing any abnormalities. The reference cluster data is created by clustering, a machine learning method. For example, the k-means method is used as a clustering algorithm. The k-means method is a clustering algorithm that classifies data into a pre-specified number of clusters. The clustering algorithm is not limited to the k-means method.

[0029] By performing clustering, data at each time included in the data showing the transition of differential pressure can be classified into clusters of data with similar characteristics. Figure 4 is a graph showing an example of clustering data on differential pressure transitions at multiple times into three clusters using the k-means method, using two explanatory variables included in the data on differential pressure transitions. In the example shown in Figure 4, the differential pressure ΔPp of the first hydraulic pressure Pp and the differential pressure ΔPp of the second hydraulic pressure Ps are each used as explanatory variables. In Figure 4, each piece of differential pressure data at multiple times is represented by a white symbol.

[0030] In Figure 4, these differential pressure data are shown in a two-dimensional space with the differential pressure ΔPp and the differential pressure ΔPs as the coordinate axes. Figure 4 shows an example of differential pressure data at multiple times being clustered into three clusters. In Figure 4, the coordinates of data classified into the first cluster are shown with open circles. In Figure 4, the coordinates of data classified into the second cluster are shown with open squares. In Figure 4, the coordinates of data classified into the third cluster are shown with open triangles. Furthermore, in Figure 4, the center of gravity of each cluster is shown with a cross. Center of gravity C_1 is the center of gravity of the first cluster. Center of gravity C_2 is the center of gravity of the second cluster. Center of gravity C_3 is the center of gravity of the third cluster.

[0031] The reference cluster data is data of the transition of differential pressures to which labels indicating the results of clustering have been added. Specifically, the reference cluster data is created by adding a label that identifies the cluster into which the data has been classified to each piece of data represented by a white symbol in the coordinate space. The reference cluster data created in this way is stored in the abnormality diagnosis system.

[0032] The abnormality diagnosis system performs abnormality determination processing using the reference cluster data. The abnormality diagnosis system performs diagnosis using data on the transition of hydraulic pressure when the hydraulic control device 240 to be diagnosed is operated in a predetermined diagnostic pattern. The predetermined diagnostic pattern is the same as the diagnostic pattern used when a hydraulic control device 240 that is not experiencing any abnormalities is operated when creating the reference cluster data.

[0033] The flow of a series of processes for abnormality diagnosis executed by the abnormality diagnosis system will be described with reference to Fig. 5. The series of processes shown in Fig. 5 are executed by a processing circuit of the abnormality diagnosis system. 5, the processing circuit of the abnormality diagnosis system first measures the oil pressure while operating the hydraulic control device 240 in a predetermined diagnostic pattern in the process of step S100. As in the case of creating the reference cluster data, the processing circuit of the abnormality diagnosis system measures the line pressure PL, the first oil pressure Pp, and the second oil pressure Ps while operating the hydraulic control device 240 in accordance with the predetermined diagnostic pattern. The processing circuit of the abnormality diagnosis system records the measured oil pressure data in the storage device.

[0034] In the process of step S110, the processing circuit of the abnormality diagnosis system creates evaluation data from data indicating the transition of the differential pressure between the oil pressure recorded in the storage device and the target oil pressure. The processing circuit of the abnormality diagnosis system records the created evaluation data in the storage device.

[0035] In the process of step S120, the processing circuit of the abnormality diagnosis system collates the evaluation data with the reference cluster data and clusters the evaluation data. Specifically, the processing circuit of the abnormality diagnosis system collates the differential pressure data at multiple times included in the evaluation data with the centroids of each cluster in the reference cluster data, and determines to which cluster each piece of data at multiple times in the evaluation data belongs.

[0036] For example, the processing circuit of the abnormality diagnosis system calculates the distance between each piece of differential pressure data at multiple times in the evaluation data and the center of gravity of each cluster, and then determines that each piece of data belongs to the cluster with the center of gravity that is closest to it.

[0037] In FIG. 4, the coordinates of one piece of differential pressure data at multiple times in the evaluation data are indicated by a black square symbol. In the example shown in FIG. 4, the centroid closest to the coordinates of this data is centroid C_2. Therefore, in this example, the processing circuit 110 determines that this data belongs to the second cluster. If the differential pressure data is significantly far from the centroid of any cluster, the processing circuit of the abnormality diagnosis system determines that the data does not belong to any cluster. In this way, the processing circuit of the abnormality diagnosis system determines to which cluster each piece of differential pressure data at multiple times included in the evaluation data belongs. The processing circuit of the abnormality diagnosis system creates evaluation cluster data by assigning labels to each piece of differential pressure data at multiple times included in the evaluation data, identifying the cluster to which the data is classified. In this way, the evaluation cluster data is data obtained by assigning labels to the differential pressure transition data for the hydraulic control device 240 to be diagnosed. The processing circuit of the abnormality diagnosis system stores the evaluation cluster data, which is the result of comparison with the reference cluster data, in a storage device.

[0038] In the process of step S140, the processing circuit of the abnormality diagnosis system executes an abnormality determination process, in which the processing circuit of the abnormality diagnosis system determines that an abnormality has occurred in the hydraulic control device 240 based on the fact that the cluster determination results between the reference cluster data and the evaluation cluster data are different.

[0039] Specifically, the processing circuit of the anomaly diagnosis system compares the reference cluster data with the evaluation cluster data. As shown in FIG. 6, the processing circuit of the anomaly diagnosis system compares the labels assigned to the data at each time in the evaluation cluster data with the labels assigned to the data at each time in the reference cluster data. In FIG. 6, label information in the reference cluster data is displayed in the "Reference Data" column, and label information in the evaluation cluster data is displayed in the "Evaluation Data" column. In this comparison, labels corresponding to the same time are compared. The processing circuit of the anomaly diagnosis system then identifies areas where the labels assigned to the evaluation cluster data do not match the labels assigned to the reference cluster data. In FIG. 6, the label of the first cluster is shown as "1," the label of the second cluster is shown as "2," and the label of the third cluster is shown as "3." In FIG. 6, if the cluster does not belong to any of these clusters, the label is shown as "-."

[0040] In the example of Figure 6, the labels at time T11 do not match. The processing circuit of the abnormality diagnosis system stores and records the comparison results in a storage device. The processing circuit of the abnormality diagnosis system then identifies the location of the abnormality based on the time when the labels did not match.

[0041] When an abnormality occurs in the hydraulic control device 240, the transition of the hydraulic pressure fluctuation when operated according to the predetermined diagnostic pattern deviates from the transition of the fluctuation when no abnormality occurs. Therefore, the labels in the evaluation cluster data, which are the result of clustering, also differ from the labels in the reference cluster data.

[0042] As described above, the predetermined diagnostic pattern is defined so that abnormalities in each location of the hydraulic control device 240 can be diagnosed from the transition of the measured hydraulic pressures by varying the line pressure PL, the first hydraulic pressure Pp, and the second hydraulic pressure Ps over a predetermined period. In other words, the predetermined diagnostic pattern is defined so that it can be diagnosed at which location an abnormality has occurred based on information about the time when a difference occurs between the label in the evaluation cluster data and the label in the reference cluster data.

[0043] For example, if an abnormality occurs in the first hydraulic damper 245, the label in the reference cluster data and the label in the evaluation cluster data will not match at time T11.

[0044] 7 shows, for each type of anomaly, the time at which the labels in the reference cluster data and the labels in the evaluation cluster data become mismatched. In FIG. 7, white circles are displayed at the times at which the labels in the reference cluster data and the labels in the evaluation cluster data match. Black circles are displayed at the times at which the labels in the reference cluster data and the labels in the evaluation cluster data become mismatched. The storage device of the anomaly diagnosis system stores information for diagnosing these types of anomalies.

[0045] The example shown in Fig. 7 indicates that when an abnormality occurs in the second hydraulic damper 246, the labels in the reference cluster data and the labels in the evaluation cluster data do not match at time T23. The example shown in Fig. 7 indicates that when an abnormality occurs in the first control valve 247, the labels in the reference cluster data and the labels in the evaluation cluster data do not match at time T41. The example shown in Fig. 7 indicates that when an abnormality occurs in the second control valve 248, the labels in the reference cluster data and the labels in the evaluation cluster data do not match at time T25 and time T45.

[0046] If no abnormality occurs in the hydraulic control device 240 to be diagnosed, the transition of hydraulic pressure fluctuations when operated according to a predetermined diagnostic pattern will not differ significantly from the transition of fluctuations when the reference cluster data was created. Therefore, the labels in the evaluation cluster data, which are the result of clustering, will match the labels in the reference cluster data. Therefore, if no abnormality occurs in any location, the labels in the reference cluster data will match the labels in the evaluation cluster data at any time.

[0047] 7 is stored in the storage device of the abnormality diagnosis system as information for diagnosing the types of abnormalities described above. In the abnormality determination process in step S140, the processing circuit of the abnormality diagnosis system refers to this information stored in the storage device to diagnose the presence or absence of an abnormality and the type of abnormality.

[0048] In the process of step S150, the processing circuit of the abnormality diagnosis system outputs the diagnosis result obtained through the abnormality determination process. Specifically, the processing circuit of the abnormality diagnosis system displays the diagnosis result on the display 130.

[0049] <Search for diagnostic patterns by the information processing device 100> Next, with reference to FIG. 8, a flow of a series of processes executed by the information processing device 100 to search for and determine a diagnostic pattern will be described.

[0050] This series of processes is executed by the processing circuitry 110 of the information processing device 100. 8, when this series of processes starts, the processing circuit 110 first sets a diagnostic pattern in the process of step S200. Each time the process of step S200 is executed, the processing circuit 110 sets the diagnostic pattern by changing the fluctuation pattern of the target hydraulic pressure PLt of the line pressure PL, the fluctuation pattern of the target hydraulic pressure Ppt of the first hydraulic pressure Pp, and the fluctuation pattern of the target hydraulic pressure Pst of the second hydraulic pressure Ps. The initial setting of the diagnostic pattern is determined in advance by the operator.

[0051] In the processing of step S210, the processing circuit 110 uses the simulation device 200 to simulate a state in which the hydraulic control device 240 is operated according to the diagnostic pattern set through the processing of step S200. Then, based on the results of the simulation, the processing circuit 110 acquires data on the line pressure PL, data on the first hydraulic pressure Pp, and data on the second hydraulic pressure Ps. The processing circuit 110 creates evaluation data in the same manner as in the processing of step S110 in FIG. 5. The processing circuit 110 creates evaluation cluster data in the same manner as in the processing of step S120 in FIG. For example, in the processing of step S210, the processing circuit 110 changes the settings of the model M_240 and performs simulations for each of the following five states:

[0052] A state in which no abnormality occurs in the hydraulic control device 240. An abnormality occurs in the first hydraulic damper 245. An abnormality occurs in the second hydraulic damper 246.

[0053] A state in which an abnormality occurs in the first control valve 247. A state in which an abnormality occurs in the second control valve 248. In the process of step S220, the processing circuit 110 executes the abnormality determination process similar to the process of step S140 in FIG.

[0054] In the process of step S230, the processing circuit 110 evaluates the diagnostic pattern based on the result of the abnormality determination process. For example, the processing circuit 110 evaluates the diagnostic pattern on the condition that all of the above five states are correctly determined.

[0055] In the process of step S240, the processing circuit 110 determines whether the result of the evaluation in step S230 indicates that the above requirements are satisfied. If the diagnostic pattern does not satisfy the requirements (step S240: NO), the processing circuit 110 returns the process to step S200. In this case, the processing circuit 110 executes the processes of steps S200 to S230 again. That is, in this case, the settings of the diagnostic pattern are changed and an attempt to execute the simulation is repeated.

[0056] If the diagnostic pattern satisfies the requirements (step S240: YES), processing circuit 110 proceeds to step S250. In the process of step S250, the processing circuit 110 records the diagnostic pattern at that time in the storage device 120, and ends this series of processes.

[0057] <Operation of this embodiment> In this way, the information processing device 100 uses the simulation device 200 that stores the model M_240 that can simulate the operation of the hydraulic control device 240 to repeatedly simulate the process of diagnosing an abnormality in the hydraulic control device 240. In this way, the information processing device 100 searches for a diagnostic pattern that is suitable for the process of diagnosing an abnormality by repeating the simulation using the simulation device 200 while changing the settings of the diagnostic pattern.

[0058] <Effects of this embodiment> (1) The information processing device 100 can easily perform a search under various settings, rather than searching for a diagnostic pattern by actually repeatedly operating the hydraulic control device 240. Therefore, the information processing device 100 can easily find a diagnostic pattern suitable for diagnosing an abnormality.

[0059] (2) By using the simulation device 200, it is possible to simulate the operation of the hydraulic control device 240 in various states. <Example of change> This embodiment can be modified as follows: This embodiment and the following modifications can be combined and implemented within the scope of technical compatibility.

[0060] The information processing device 100 may be configured to search for a diagnostic pattern using the simulation device 200. For example, an operator may evaluate the diagnostic patterns that have been tried and ultimately decide on a diagnostic pattern. For example, an operator may set the diagnostic pattern.

[0061] The abnormality diagnosis system does not need to determine the type of abnormality. For example, it outputs the presence or absence of an abnormality and label information for each time shown in Figure 7. In this case, as in the above embodiment, the diagnosis pattern may be one in which time periods are divided according to the type of abnormality to be detected, and the hydraulic pressures controlled by the hydraulic control device 240 are controlled in sequence in a pattern that makes it easy to detect each abnormality.

[0062] In this case, the abnormality diagnosis system outputs information indicating the time period during which the reference cluster data and the evaluation cluster data diverge, along with information on the determination result as to whether or not an abnormality has occurred in the hydraulic control device 240. In this case, the abnormality diagnosis system can provide information that serves as material for determining not only whether or not an abnormality has occurred, but also the type of abnormality that has occurred.

[0063] The abnormality diagnosis system determines whether or not there is an abnormality in the hydraulic control device 240 that controls the hydraulic pressure supplied to the first pulley 210 and the second pulley 220 of the continuously variable transmission. The hydraulic control device 240 to be diagnosed is not limited to the hydraulic control device 240 of a continuously variable transmission. The diagnosis target may also be a hydraulic control device of a stepped transmission. Furthermore, the diagnosis target is not limited to a hydraulic control device of an automatic transmission.

[0064] In the above embodiment, the information processing device 100 includes a processing circuit 110 and a storage device 120 and executes software processing. However, this is merely an example. For example, the information processing device 100 may include a dedicated hardware circuit (e.g., an ASIC) that processes at least part of the software processing executed in the above embodiment. That is, the information processing device 100 may have any of the following configurations (A) to (C): (A) The information processing device 100 includes an execution device that executes all processing in accordance with a program and a storage device that stores the program. That is, the information processing device 100 includes a software execution device. (B) The information processing device 100 includes an execution device that executes part of the processing in accordance with a program and a storage device. Furthermore, the information processing device 100 includes a dedicated hardware circuit that executes the remaining processing. (C) The information processing device 100 includes a dedicated hardware circuit that executes all processing. Here, there may be multiple software execution devices and / or dedicated hardware circuits. That is, the above processing may be executed by a processing circuitry that includes at least one of one or more software execution devices and one or more dedicated hardware circuits. The storage devices or computer-readable media for storing the programs include any available media that can be accessed by a general purpose or special purpose computer. [Explanation of symbols]

[0065] 100...information processing device, 110...processing circuit, 120...storage device, 130...display, 200...simulation device, M_240...model, 210...first pulley, 220...second pulley, 240...hydraulic control device, 241...regulator valve, 242...modulator valve, 243...first linear solenoid valve, 244...second linear solenoid valve, 245...first hydraulic damper, 246...second hydraulic damper, 247...first control valve, 248...second control valve

Claims

1. The storage device stores reference cluster data obtained by classifying differential pressure data at multiple times included in data on the transition of the differential pressure between the hydraulic pressure and the target hydraulic pressure when a hydraulic control device in which no abnormality has occurred is operated in a predetermined diagnostic pattern into a predetermined number of clusters by clustering, which is machine learning; The processing circuit creating, as evaluation data, data indicating a transition of a differential pressure between a target hydraulic pressure and a hydraulic pressure recorded while the hydraulic control device to be evaluated is operated in accordance with the diagnostic pattern; creating evaluation cluster data by comparing differential pressure data at a plurality of times included in the evaluation data with the center of gravity of each cluster in the reference cluster data, and determining to which of the predetermined number of clusters in the reference cluster data each piece of data at a plurality of times in the evaluation data belongs; an information processing device that determines the diagnostic pattern to be used in an abnormality diagnosis system that determines that an abnormality has occurred in the hydraulic control device based on a deviation between a determination result of the cluster between the reference cluster data and the evaluation cluster data, A simulation device that stores a model of the hydraulic control device and is capable of simulating the operation of the hydraulic control device is used to repeatedly simulate a process for diagnosing an abnormality in the hydraulic control device, thereby searching for the diagnostic pattern that is suitable for the process for diagnosing the abnormality. Information processing device.

2. The diagnostic pattern is a pattern in which a time period is divided according to the type of abnormality to be determined, and a plurality of hydraulic pressures controlled by the hydraulic control device are controlled in order in a pattern that makes it easy to determine each abnormality, In each time period, the diagnostic pattern that generates differential pressure data corresponding to the corresponding type of abnormality is searched for. The information processing device according to claim 1 .

3. The diagnostic pattern is searched for to determine whether or not there is an abnormality in a hydraulic control device that controls hydraulic pressure supplied to a first pulley and a second pulley in a continuously variable transmission. The information processing device according to claim 1 .

Images