Information processing device

Abstract

To provide an information processing device that makes it possible to avoid vehicle failures by a method other than component replacement.SOLUTION: Stored in an information processing device 200 are a failure prediction model 140 for outputting the type of failure which is predicted to occur, and a simulation model 240 for outputting the time-series data of behavior of each vehicle part when the time-series data of driving operation is inputted. When the data of input variable at the time when a predicted type of failure is outputted from the failure prediction device 100 is acquired, the information processing device 200 calculates the simulated data of the time-series data of behavior of each vehicle part for the case where the value of control parameter of a vehicle 400 is changed by inputting the time-series data of driving operation to the simulation model 240. The information processing device 200 inputs an input variable for trial composed of the time-series data of driving operation and the simulated data to the failure prediction model 140 and searches for the value of a control parameter with which failure occurrence is no longer predicted, and presents a change to the obtained value to the user of the vehicle 400.SELECTED DRAWING: Figure 1

Description

Technical Field

[0005] ,

[0001] This invention relates to an information processing apparatus.

Background Art

[0002] Patent Document 1 discloses a failure prediction device. This failure prediction device generates a learning model for failure prediction using time-series sensor data until the device to be monitored fails as teacher data. The failure prediction device predicts the probability that the device to be monitored fails by using the generated learning model and taking the time-series sensor data received from the device to be monitored as input.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] When a failure prediction device equipped with a failure prediction model learned by such supervised learning is applied to a vehicle and the failure of a vehicle part is predicted, the user of the vehicle has to replace the part in order to avoid the occurrence of the failure.

Means for Solving the Problems

[0005] ​​The information processing device for solving the above problems comprises a processing unit and a storage unit. In the information processing unit, a failure prediction model, learned through supervised learning, is stored in the storage unit to output the type of failure that is predicted to occur when input variables consisting of time-series data of driving operations and the behavior of various parts of the vehicle are input. In addition, the information processing unit is stored in the storage unit a simulation model that outputs time-series data of the behavior of various parts of the vehicle when time-series data of driving operations is input. In this information processing device, the processing unit obtains the data of the input variables when the failure prediction device outputs the type of failure that is predicted by the failure prediction device, which predicts vehicle failures using the same failure prediction model as the processing unit. In this information processing device, the device inputs the time-series data of the driving operations in the acquired input variable data into the simulation model and calculates simulated data that simulates the time-series data of the behavior of each part of the vehicle when the values ​​of the vehicle's control parameters are changed. This generates a trial input variable consisting of the time-series data of the driving operations in the acquired input variable data and the simulated data. The generated trial input variable is then input into the failure prediction model to search for a control parameter value that will no longer cause a failure. In this information processing device, the device presents the vehicle user with a change to the control parameter to a value obtained through the search that will no longer cause a failure.

[0006] A second information processing device for solving the above problems is connected to the vehicle via a communication network. This information processing device comprises a processing unit and a storage device. In this information processing device, a failure prediction model, learned through supervised learning, is stored in the storage device to output the type of failure that is predicted to occur when input variables consisting of time-series data of driving operations and the behavior of various parts of the vehicle are input. In addition, in this information processing device, a simulation model that outputs time-series data of the behavior of various parts of the vehicle when time-series data of driving operations is input is stored in the storage device. In this information processing device, the processing unit receives the input variable data from the vehicle and inputs it into the failure prediction model to output the type of failure that is predicted to occur. In this information processing device, the device inputs the time-series data of the driving operations in the input variable data when it outputs the type of failure that is predicted to occur into the simulation model, and calculates simulated data that simulates the time-series data of the behavior of each part of the vehicle when the value of the vehicle's control parameter is changed. This generates a trial input variable consisting of the time-series data of the driving operations in the acquired input variable data and the simulated data, and inputs the generated trial input variable into the failure prediction model to search for a value of the control parameter that will no longer cause a failure to occur. In this information processing device, the device presents the user of the vehicle with a change to the control parameter to a value obtained through the search that will no longer cause a failure to occur. [Effects of the Invention]

[0007] These information processing devices will allow users to avoid vehicle malfunctions through methods other than replacing parts. [Brief explanation of the drawing]

[0008] [Figure 1] Figure 1 is a schematic diagram showing the configuration of a fault avoidance system, including an information processing device according to one embodiment. [Figure 2]Figure 2 is a sequence diagram showing the processing flow by the vehicle control device, the fault prediction device, and the information processing device in the fault avoidance system of the embodiment. [Figure 3] Figure 3 is a flowchart showing the processing flow executed by the information processing device in the fault avoidance system of the embodiment. [Figure 4] Figure 4 is a flowchart showing the processing flow related to vehicle failure prediction performed by the failure prediction device in the failure avoidance system of the embodiment. [Figure 5] Figure 5 is a flowchart showing the processing flow related to the presentation of control parameters performed by the fault prediction device in the fault avoidance system of the embodiment. [Figure 6] Figure 6 is a schematic diagram showing the configuration of a fault avoidance system, including an example of a modified information processing device. [Modes for carrying out the invention]

[0009] An embodiment of the information processing device will be described below with reference to Figures 1 to 5. <Configuration of the failure avoidance system 10> As shown in Figure 1, the fault avoidance system 10 consists of an information processing device 200 and a vehicle 400. The information processing device 200 is connected to the vehicle 400 via a communication network 300 so as to be able to communicate with it.

[0010] Vehicle 400 is equipped with a fault prediction device 100, a vehicle control device 410, and a display device 420. The fault prediction device 100 includes a storage device 120 in which a program and a fault prediction model 140 are stored, and a processing device 110 that executes the program stored in the storage device 120 and performs various processes. The processing device 110 includes a processor.

[0011] The failure prediction model 140 is a model trained through supervised learning to output the type of failure that is predicted to occur when input variables consisting of time-series data of driving operations and the behavior of various parts of the vehicle are input. The time-series data of driving operations refers to data that shows the amount of operation performed by the user on the vehicle 400, such as accelerator opening, brake operation amount, and steering operation amount. The time-series data of the behavior of various parts of the vehicle refers to data that shows the state of various parts of the vehicle 400, such as engine speed, vehicle speed, engine coolant temperature, boost pressure, and intake air volume.

[0012] The fault prediction device 100 includes a communication device 130. The communication device 130 is implemented as hardware such as a network adapter, various communication software, or a combination thereof. The communication device 130 is configured to enable wireless communication with the information processing device 200 via the communication network 300. The fault prediction device 100 is also connected to the vehicle control device 410 and the display device 420 to exchange information.

[0013] The vehicle control device 410 controls various parts of the vehicle 400. Various sensors are connected to the vehicle control device 410 to detect the state of the vehicle 400. The vehicle control device 410 can acquire data of the aforementioned input variables in the vehicle 400 from the various sensors.

[0014] The display device 420 comprises an input device for the user to input information and a display device for displaying information to the user. The display device 420 displays information received from the fault prediction device 100 to the user of the vehicle 400. The display device 420 is, for example, a car navigation system installed in the vehicle 400. The display device 420 may also constitute a part of the fault prediction device 100 as a display unit.

[0015] The information processing device 200 includes a processing device 210, a storage device 220, and a communication device 230. The information processing device 200 can be configured using a plurality of computers. For example, the information processing device 200 can be configured by a plurality of server devices.

[0016] The storage device 220 stores a program, a failure prediction model 140, and a simulation model 240. The failure prediction model 140 stored in the storage device 220 is the same model as the failure prediction device 100. The simulation model 240 outputs time-series data of the behavior of each part of the vehicle when input with time-series data of driving operations.

[0017] The processing device 210 executes a program stored in the storage device 220 to execute various processes. Note that the processing device 110 includes a processor. The communication device 230 is implemented as hardware such as a network adapter, various communication software, or a combination thereof. And the communication device 230 is configured to enable wireless communication with the failure prediction device 100 via the communication network 300.

[0018] <Transmission and reception of data in the failure avoidance system 10> Next, the transmission and reception of data among the failure prediction device 100, the information processing device 200, and the vehicle control device 410 will be described with reference to FIG. 2.

[0019] As shown in the upper part of FIG. 2, the vehicle control device 410 transmits data of input variables acquired from various sensors of the vehicle 400 to the failure prediction device 100. Such data transfer is performed at timing at regular intervals set for the failure prediction device 100 to perform failure prediction on the vehicle 400 regularly.

[0020] Upon receiving the input variable data, the fault prediction device 100 inputs the received input variables into the fault prediction model 140 and outputs the types of faults predicted to occur for the vehicle 400. If the probability of any type of fault occurring is low, the fault prediction device 100 does not output any type of fault. In other words, in this case, the fault prediction device 100 does not predict the occurrence of a fault.

[0021] As shown in the upper part of Figure 2, when the fault prediction device 100 predicts the occurrence of any fault, it transmits information about the type of fault predicted and the input variable data at the time the fault type was output to the information processing device 200.

[0022] The information processing device 200 then uses the failure prediction model 140 to search for control parameter values ​​that will prevent failures in the vehicle 400. As shown in the lower part of Figure 2, if the information processing device 200 finds a control parameter value that can avoid failures through its search, it transmits that control parameter value to the failure prediction device 100.

[0023] Subsequently, the fault prediction device 100 presents the user of the vehicle 400 with a change to the control parameter of the vehicle 400, based on the value received from the information processing device 200. As shown in the lower part of Figure 2, if the user approves the change to the presented value, the fault prediction device 100 transmits the value of the control parameter to the vehicle control device 410. The vehicle control device 410 changes the control parameter of the vehicle 400 based on the received value. In this way, the fault avoidance system 10 proposes changes to the control parameter of the vehicle 400 so that the occurrence of a fault is no longer predicted.

[0024] The following will provide a more detailed explanation of the processing flow executed in the information processing device 200 and the processing flow executed in the fault prediction device 100 in order to implement the proposed changes to the control parameters as described above, with reference to Figures 3 to 5.

[0025] <Processing performed by the information processing device 200> Figure 3 shows the flow of a series of processes executed in the information processing device 200. This series of processes is executed by the processing unit 210 of the information processing device 200. This series of processes corresponds to the processes executed by the information processing device 200 when it receives information on the type of fault and input variable data from the fault prediction device 100, as shown at the top of Figure 2.

[0026] As shown in Figure 3, when this series of processes is started, the processing unit 210 first performs an extraction process in step S100. The extraction process narrows down the types of control parameters in the vehicle 400 that will be changed during the search process described later, based on the fault type information received from the fault prediction device 100.

[0027] The storage device 220 stores a database containing the types of control parameters to be changed during the search process, corresponding to the types of failures predicted by the failure prediction device 100. In the extraction process, the processing device 210 extracts the types of control parameters to be changed during the search process from the database based on the information about the type of failure. This allows the processing device 210 to narrow down the types of control parameters to be changed during the search process. The range of the search during the search process is predetermined according to the types of control parameters to be changed. Once the extraction process is complete, the process proceeds to the next step S110.

[0028] In step S110, the processing unit 210 starts a search process. The search process searches for a control parameter value that will cause the failure prediction model 140 to no longer output the type of failure predicted to occur in the vehicle 400. From here on, the series of processes from step S120 to step S160 are all performed as part of the search process. In the search process, the processing unit 210 repeats the processes from step S120 to step S160 and terminates the search process when the condition in step S170 is met.

[0029] In step S120, the processing unit 210 determines one type of control parameter and its modified value to be changed in the simulation process of the next step, S130. The type of control parameter to be changed is determined from the types narrowed down in the extraction process. The modified value of the determined control parameter is determined from a predetermined range.

[0030] In step S130, the processing unit 210 performs simulation processing. Simulation processing is the process of generating trial input variables. Trial input variables are input variables to be input to the failure prediction model 140 in the failure prediction processing of step S140.

[0031] In the simulation process, the processing unit 210 inputs time-series data of driving operations from the input variables received from the fault prediction device 100 into the simulation model 240. The processing unit 210 then uses the simulation model 240 to calculate simulated data, which is time-series data of the behavior of each part of the vehicle 400 when the control parameter changes determined in step S120 are implemented. Subsequently, the processing unit 210 generates trial input variables by combining the time-series data of driving operations input into the simulation model 240 and the simulated data calculated by the simulation model 240.

[0032] In step S140, the processing unit 210 performs failure prediction processing. In the failure prediction processing, the processing unit 210 inputs trial input variables into the failure prediction model 140. In step S150, the processing unit 210 determines whether a failure in the vehicle 400 is predicted based on the failure prediction process in step S140. Specifically, in step S140, the processing unit 210 inputs trial input variables into the failure prediction model 140 and determines whether the type of failure predicted for the vehicle 400 is output. If the failure prediction model 140 outputs the type of failure predicted for the vehicle 400, the processing unit 210 determines that a failure has been predicted. On the other hand, if the failure prediction model 140 does not output the type of failure predicted for the vehicle 400, the processing unit 210 determines that a failure was not predicted.

[0033] If the processing unit 210 determines that no failure was predicted during the process in step S150 (step S150: NO), the process proceeds to step S160. In the next step, S160, the processing unit 210 stores the control parameter values ​​determined in step S120 and the trial input variable data output in the simulation process of step S130 in the storage device 220. By performing the process in step S160, the information processing unit 200 can determine the control parameter values ​​for which the occurrence of a failure in the vehicle 400 is no longer predicted. When the process in step S160 is completed, the processing unit 210 returns to step S120 without moving the process to step S170.

[0034] If the processing unit 210 determines that a failure is predicted in step S150 (step S150: YES), the processing unit 210 does not perform the processing in step S160. In this case, the processing unit 210 returns to step S120. In this way, the series of processes from step S120 to step S160 is repeated regardless of whether a failure was predicted in step S150.

[0035] If the processes in steps S120 to S160 have been performed for all the control parameters narrowed down in step S100 across the entire search range, the process moves on to the next step, S170. In other words, the processes in steps S120 to S160 are repeated until the search for all the control parameters narrowed down in the extraction process has been completed across the entire search range. When the process moves to step S170, the search process ends.

[0036] In the next step, S180, the processing unit 210 performs a control parameter selection process. The control parameter selection process is the process of selecting a value to present to the user from among the control parameter values ​​that the storage device 220 has stored in step S160 for which the occurrence of a failure is no longer predicted for the vehicle 400.

[0037] The processing unit 210 selects one value from among multiple control parameter values ​​stored in the memory device 220 based on the trial input variables stored in the memory device 220 along with the control parameter values ​​in step S160. The processing unit 210 evaluates the impact on the vehicle 400's dynamic performance when the control parameter values ​​are changed, based on the simulated data calculated in the simulation process of step S130 from among the trial input variables. The vehicle 400's dynamic performance refers to the performance related to the vehicle 400's driving, such as acceleration, braking, and turning. The processing unit 210 selects the control parameter value that is evaluated as having the least adverse impact on the vehicle 400's driving performance from among the multiple control parameter values ​​to present to the user. If there is only one control parameter value stored in step S160 in the search process that is unlikely to cause a failure, that value is selected to present to the user.

[0038] In step S190, the processing unit 210 determines whether or not there is a suitable control parameter value to present to the user. In step S190, if a control parameter value to present to the user has been selected through step S180, the processing unit 210 determines that there is a suitable control parameter value. On the other hand, it may not be possible to obtain a control parameter value that would prevent a failure from occurring for the vehicle 400 through the search process from steps S110 to S170. In this case, in step S190, the processing unit 210 determines that there is no suitable control parameter value.

[0039] In step S190, if the processing unit 210 determines that there is an appropriate value for the control parameter (step S190: YES), the process proceeds to step S200. In step S200, the processing unit 210 transmits data indicating the value of the control parameter selected in step S180 to the fault prediction device 100. After transmitting the data indicating the value of the control parameter, the processing unit 210 terminates this series of processes.

[0040] If the processing unit 210 determines in step S190 that there is no appropriate value for the control parameter (step S190: NO), the process proceeds to step S210. In step S210, the processing unit 210 sends a signal to the fault prediction device 100 indicating that there is no appropriate value for the control parameter. If the signal is sent, the processing unit 210 terminates this series of processes.

[0041] <An example of processing performed by the information processing device 200> The following provides a specific example of the processing performed by the information processing device 200, as explained in Figure 3. In this example, the user is increasing the boost pressure by rewriting the boost pressure control parameters set in the vehicle control device 410 in order to enjoy driving on a circuit. Here, we present a case where, under such circumstances, the fault prediction device 100 outputs that the turbocharger turbine blades are damaged. At this time, as shown at the top of Figure 2, the processing device 210 of the information processing device 200 receives information from the fault prediction device 100 indicating that the turbocharger turbine blades are damaged, as well as data of the input variables that were input to the fault prediction model 140 when the damage was output.

[0042] The processing unit 210, upon receiving the data, performs the extraction process in step S100. The processing unit 210 extracts from the database stored in the storage device 220 the types of control parameters to be changed during the search process in response to damage to the turbocharger's turbine blades. Examples of such control parameters include, for example, a control parameter for boost pressure, since reducing the boost pressure reduces the load on the turbocharger's turbine blades. Another example of such control parameters is a control parameter for throttle opening, since reducing the throttle opening reduces the load on the turbocharger's turbine blades. Yet another example of such control parameters is a control parameter for fuel injection amount, since increasing the fuel injection amount lowers the temperature inside the combustion chamber and reduces the load on the turbocharger's turbine blades.

[0043] After the extraction process in step S100 is completed, the processing unit 210 performs the search process in steps S110 to S170 for control parameters related to boost pressure, fuel injection amount, and throttle opening, within predetermined ranges.

[0044] In the search process, if values ​​for control parameters such as boost pressure, fuel injection amount, and throttle opening are obtained in which the occurrence of a failure is no longer predicted, the processing unit 210 performs the control parameter selection process in step S180. The processing unit 210 evaluates the impact of each obtained control parameter on the vehicle 400's dynamic performance. For example, if changing the control parameter to increase the fuel injection amount has the least adverse effect on dynamic performance, then a value that decreases the fuel injection amount is selected as the control parameter value to present to the user. After performing the selection process, the processing unit 210 transmits data indicating the selected control parameter value in step S200 to the fault prediction device 100.

[0045] In this way, the information processing device 200 searches for control parameter values ​​that can avoid damage to the turbine blades and presents them to the user. <Processing when the fault prediction device 100 predicts a fault in the vehicle 400> Figure 4 shows the flow of a series of processes performed when the fault prediction device 100 predicts a fault in the vehicle 400. This series of processes is performed by the processing unit 110 of the fault prediction device 100. This series of processes corresponds to the processes performed by the processing unit 110 of the fault prediction device 100 when the fault prediction device 100 receives input variable data from the vehicle control device 410, as shown at the top of Figure 2.

[0046] As shown in Figure 4, when this series of processes is started, the processing unit 110 first performs a fault prediction process in step S300. In the fault prediction process, the processing unit 110 takes the input variable data received from the vehicle control device 410 as input and outputs the type of fault that is predicted to occur using the fault prediction model 140 stored in the storage device 120.

[0047] In step S310, the processing unit 110 determines whether a failure in the vehicle 400 is predicted as a result of the processing in step S300. Specifically, the processing unit 110 inputs the input variables received from the vehicle control device 410 in step S300 into the failure prediction model 140. The processing unit 110 then determines whether the type of failure predicted for the vehicle 400 is output. If the failure prediction model 140 outputs the type of failure predicted for the vehicle 400, the processing unit 110 determines that a failure in the vehicle 400 is predicted. On the other hand, if the failure prediction model 140 does not output the type of failure predicted for the vehicle 400, the processing unit 110 determines that a failure in the vehicle 400 is not predicted.

[0048] If the processing unit 110 determines in step S310 that a malfunction in the vehicle 400 is predicted (step S310: YES), the process proceeds to step S320. In step S320, the processing unit 110 transmits to the information processing unit 200 the information about the predicted type of malfunction output in step S300 and the input variable data at the time the type of malfunction was output. After completing the processing in step S320, the processing unit 110 terminates this series of processes.

[0049] If the processing unit 110 determines that no malfunction occurred in the vehicle 400 during the process of step S310 (step S310: NO), the processing unit 110 terminates this series of processes. In other words, the processing unit 110 terminates this series of processes without performing the process of step S320.

[0050] <Processing performed by the fault prediction device 100 after receiving information from the information processing device 200> Figure 5 shows the sequence of processes executed when the fault prediction device 100 receives information from the information processing device 200. This sequence of processes is executed by the processing device 110 of the fault prediction device 100. This sequence of processes corresponds to the processes executed when the fault prediction device 100 receives information from the information processing device 200, as shown at the bottom of Figure 2.

[0051] As shown in Figure 5, when this series of processes is started, the processing unit 110 first determines in step S400 whether it has received information from the information processing unit 200 indicating that there is no appropriate control parameter value. If the processing unit 110 determines in step S400 that it has not received a signal (step S400: NO), the process proceeds to step S410.

[0052] In step S410, the processing unit 110 determines whether it has received data indicating the value of the control parameter from the information processing unit 200. If the processing unit 110 determines that it has not received data indicating the value of the control parameter (step S410: NO), the processing of step S410 is repeated.

[0053] In step S410, if the processing unit 110 determines that it has received data indicating the value of a control parameter (step S410: YES), the process proceeds to step S420. In step S420, the processing unit 110 presents the user with the change to the control parameter received from the information processing unit 200. For example, the processing unit 110 displays on the display device 420 the type of control parameter to be changed, the changed value, and text information indicating the impact on the vehicle 400 due to the change in value.

[0054] In the next step, S430, the processing unit 110 determines whether the user has authorized the change to the control parameter presented in step S420. That is, after presenting the change to the value of the control parameter received from the information processing unit 200, the processing unit 110 asks whether the user wishes to change to the presented control parameter.

[0055] For example, the processing unit 110 displays a message on the display device 420 to ask whether the user consents to changing the value of the control parameter. If the user performs an action indicating their intention to consent to the change in response to this inquiry, the processing unit 110 determines that the user has consented to changing the value. On the other hand, if the user performs an action indicating their intention not to consent to the change, the processing unit 110 determines that the user has not consented to changing the value.

[0056] In step S430, if the processing unit 110 determines that the user has authorized the change to the value of the control parameter presented (step S430: YES), the process proceeds to step S440. In step S440, the processing unit 110 transmits the data of the presented control parameter value to the vehicle control device 410. After transmitting the data, the processing unit 110 terminates this series of processes. Upon receiving the data transmitted in step S440, the vehicle control device 410 changes the value of the control parameter to the received value.

[0057] On the other hand, if the processing unit 110 determines in step S400 that it has received a signal from the information processing unit 200 indicating that there is no appropriate value for the control parameter (step S400: YES), the process proceeds to step S450. Also, if the processing unit 110 determines in step S430 that the user has not authorized the change of the control parameter value (step S430: NO), the process proceeds to step S450. In the process of step S450, the processing unit 110 proposes to the user the replacement of the component that is expected to fail, according to the type of failure that is expected. For example, the processing unit 110 proposes the replacement of the component that is expected to fail by displaying text information on the display device 420.

[0058] <Operation of this embodiment> The information processing device 200 in the fault avoidance system 10 searches for a control parameter value that makes the occurrence of a fault unpredictable and presents the user with a change to the discovered control parameter value. In other words, the information processing device 200 presents a change in the control parameter value as a countermeasure to avoid the occurrence of a fault.

[0059] <Effects of this embodiment> (1) The failure avoidance system 10 allows the user to avoid vehicle 400 failures by means other than replacing parts.

[0060] (2) If the search yields multiple values ​​for the control parameter in which failure is no longer predicted, the processing unit 210 of the information processing unit 200 selects the value from among the obtained values ​​that has the least adverse effect on the vehicle 400's dynamic performance. The processing unit 210 then presents the user with a change to the selected value for the control parameter. Therefore, the information processing unit 200 presents the user with the value from among the discovered values ​​that has the least adverse effect on the vehicle 400's dynamic performance. In this way, the information processing unit 200 can suppress the deterioration of dynamic performance caused by changes in the control parameter values.

[0061] (3) The storage device 220 of the information processing device 200 stores a database containing the types of control parameters to be changed during the search, corresponding to the types of failures predicted by the failure prediction device 100. The processing device 210 acquires information on the types of failures predicted by the failure prediction device 100. Then, it extracts the types of control parameters to be changed during the search from the database based on the information on the predicted types of failures, thereby narrowing down the types of control parameters to be changed during the search. As a result, the information processing device 200 narrows down the types of control parameters to be changed during the search based on the information on the types of failures predicted by the failure prediction device 100. This allows the information processing device 200 to quickly find the values ​​of control parameters for which failures are no longer predicted.

[0062] <Example of changes> This embodiment can be implemented with the following modifications. This embodiment and the following modifications can be combined with each other to the extent that they do not contradict each other technically.

[0063] The configuration of the fault avoidance system 10, including the modified information processing device, will be explained with reference to Figure 6. <Configuration of the fault avoidance system 10, including the modified information processing device> Figure 6 shows the overall configuration of the failure avoidance system 10 when the information processing device 200 is configured in a different way than the configuration described with reference to Figures 1 to 5.

[0064] Unlike the fault avoidance system 10 described with reference to Figure 1, the fault avoidance system 10 employing the information processing device 200 shown in Figure 6 does not have a fault prediction device 100 installed in the vehicle 400. Instead, the information processing device 200 is equipped with a fault prediction unit 500. The fault prediction unit 500 includes a storage device 520 that stores a program and a fault prediction model 140, and a processing unit 510 that executes the program stored in the storage device 520 and performs various processes. The processing unit 510 includes a processor.

[0065] In the fault avoidance system 10 employing the information processing device 200 shown in Figure 6, the fault prediction unit 500 of the information processing device 200 acquires input variable data from the vehicle 400 and outputs the type of fault that is predicted to occur in the vehicle 400. Thus, in the fault avoidance system 10 employing the information processing device 200 shown in Figure 6, the fault prediction unit 500 of the information processing device 200 takes on the role that was played by the fault prediction device 100 in the fault avoidance system 10 explained with reference to Figure 1.

[0066] This information processing device 200 can also achieve the same effects as the information processing device 200 with the configuration described with reference to Figure 1. In the above embodiment, an example was shown in which the processing unit 110 displays on a display device 420 provided in the vehicle 400 the type of control parameter to be changed, the changed value, and text information indicating the impact on the vehicle 400 due to the change in value. An example was also shown in which the processing unit 110 displays a message on the display device 420 to request a response from the user regarding whether or not to permit the change in the value of the control parameter. The display device that performs such display does not have to be the display device 420 provided in the vehicle 400. For example, a device such as a smartphone owned by the user can be used as the display device.

[0067] This example shows how the information processing device 200 performs an extraction process to narrow down the types of control parameters to be changed during the search. The information processing device 200 does not necessarily have to perform an extraction process. For example, the types of control parameters that can be changed may be limited from the start. In such cases, there is no need to perform an extraction process to narrow down the types of control parameters to be changed during the search.

[0068] In the fault avoidance system 10 described above, the fault prediction device 100 acquires input variable data from the vehicle control device 410. Alternatively, the fault prediction device 100 may acquire input variable data directly from sensors in various parts of the vehicle 400.

[0069] In the fault avoidance system 10, which employs the configurations shown in Figures 1 and 6, the information processing device 200 is installed outside the vehicle 400 as multiple server devices. In contrast, the information processing device 200 may be mounted on the vehicle 400 and constitute part of the vehicle 400.

[0070] In the fault avoidance system 10 described above, the fault prediction device 100 presented only the value of one control parameter selected by the information processing device 200 in the control parameter selection process of step S180 in Figure 3, in step S420 in Figure 5. In contrast, the fault prediction device 100 may present multiple values ​​as the values ​​of the control parameter to be presented and allow the user to select the value of the control parameter to which the change will be applied.

[0071] In the fault avoidance system 10 described above, the fault prediction device 100 asks the user whether they authorize a change in the control parameter value, and if the user authorizes it, it changes the control parameter value in the vehicle 400. In contrast, the fault prediction device 100 may change the control parameter value without requiring the user's authorization. In this case, the processing device 210 transmits and presents a control parameter value for which a fault is no longer predicted, and causes the vehicle 400 to change the control parameter to the presented value. As a result, the information processing device 200 can prevent a fault in the vehicle 400. [Explanation of symbols]

[0072] 10...Fault avoidance system, 100...Fault prediction device, 110...Processing device, 120...Storage device, 130...Communication device, 140...Fault prediction model, 200...Information processing device, 210...Processing device, 220...Storage device, 230...Communication device, 240...Simulation model, 300...Communication network, 400...Vehicle, 410...Vehicle control device, 420...Display device, 500...Fault prediction unit, 510...Processing device, 520...Storage device

Claims

1. A device comprising a processing unit and a memory device, A fault prediction model, trained through supervised learning, takes input variables consisting of time-series data of driving operations and the behavior of various parts of the vehicle as input and outputs the type of fault that is predicted to occur. A simulation model that takes time-series data of driving operations as input and outputs time-series data of the behavior of various parts of the vehicle is stored in the memory device. The aforementioned processing apparatus To obtain the data of the input variables when the fault prediction device outputs the type of fault predicted by the fault prediction device, using the same fault prediction model as the fault prediction device that predicts vehicle failures, The time-series data of the driving operations in the acquired input variable data is input into the simulation model to calculate simulated data that simulates the time-series data of the behavior of each part of the vehicle when the values ​​of the vehicle's control parameters are changed, thereby generating trial input variables consisting of the time-series data of the driving operations in the acquired input variable data and the simulated data, and the generated trial input variables are input into the failure prediction model to search for control parameter values ​​that will no longer predict the occurrence of failures. To present the user of the vehicle with a change to the control parameter to a value obtained through exploration that makes the occurrence of a fault no longer predictable, and to perform the following: Information processing device.

2. If the search yields multiple control parameter values ​​for which failure is no longer predicted, the processing unit selects the value from among the obtained values ​​that has the least adverse effect on the vehicle's dynamic performance and presents the user with a change to the selected control parameter. The information processing apparatus according to claim 1.

3. The aforementioned processing apparatus The vehicle is instructed to change the control parameters to the specified values. The information processing apparatus according to claim 1 or 2.

4. A database containing the types of control parameters to be changed during the search process, corresponding to the types of faults predicted by the fault prediction device, is stored in the storage device. The aforementioned processing apparatus The failure prediction device acquires information on the type of failure it predicts, The process involves extracting the types of control parameters to be changed during the search from the database based on information about the predicted types of failures, thereby narrowing down the types of control parameters to be changed during the search. The information processing apparatus according to claim 1.

5. It is an information processing device connected to the vehicle via a communication network. A device comprising a processing unit and a memory device, A fault prediction model, trained through supervised learning, takes input variables consisting of time-series data of driving operations and the behavior of various parts of the vehicle as input and outputs the type of fault that is predicted to occur. A simulation model that takes time-series data of driving operations as input and outputs time-series data of the behavior of various parts of the vehicle is stored in the memory device. The aforementioned processing apparatus The system receives data of the input variables from the vehicle and inputs it into the failure prediction model to output the type of failure that is predicted to occur. The process involves inputting the time-series data of the driving operations in the input variable data when the type of failure predicted to occur is output into the simulation model, calculating simulated data that simulates the time-series data of the behavior of each part of the vehicle when the values ​​of the vehicle's control parameters are changed, thereby generating a trial input variable consisting of the time-series data of the driving operations in the acquired input variable data and the simulated data, and inputting the generated trial input variable into the failure prediction model to search for a control parameter value that will no longer predict the occurrence of a failure. To present the user of the vehicle with a change to the control parameter to a value obtained through exploration that makes the occurrence of a fault no longer predictable, and to perform the following: Information processing device.

Images