Vehicle Evaluation System

Abstract

To provide a vehicle evaluation system capable of performing evaluation by determining a difference in a degree of a vehicle from sound data.SOLUTION: A vehicle evaluation system includes a processing circuit 110 and a storage device 120. Data of a learned model trained so as to generate reference operation data from training sound data are stored in the storage device 120. The processing circuit 110 executes generation processing for outputting generation data as the reference operation data generated by inputting evaluation sound data collected by operating an object vehicle 10 for predetermined time to the learned model. The processing circuit 110 also executes evaluation processing for evaluating the object vehicle 10 according to a deviation level by comparing object operation data for showing an operation situation when the object vehicle 10 is operated for predetermined time with the generation data.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a vehicle evaluation system for evaluating vehicles. [Background technology]

[0002] The abnormal sound detection device disclosed in Patent Document 1 detects the occurrence of an abnormal sound using the frequency spectrum of measured sound data. Specifically, the abnormal sound detection device calculates the area exceeding a threshold level in the frequency spectrum of the measured sound data. The abnormal sound detection device then compares the calculated area with a determination value to determine whether an abnormal sound has occurred. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2014-222189 A Summary of the Invention [Problem to be solved by the invention]

[0004] There is a possibility of an evaluation system that evaluates a vehicle by analyzing sound data recorded from the vehicle. However, in order to evaluate a vehicle, it is necessary not only to distinguish from the sound data whether the vehicle is obviously broken and making an abnormal sound, but also to distinguish the difference in the level of the vehicle from the sound data and evaluate it. Therefore, a vehicle evaluation system suitable for evaluating vehicles is required. [Means for solving the problem]

[0005] The means for solving the above problems and their effects will be described below. A vehicle evaluation system for solving the above problem is a vehicle evaluation system that evaluates a target vehicle using sound data recorded from a target vehicle that is a vehicle to be evaluated. The vehicle evaluation system includes a processing circuit and a storage device. The storage device stores data of a learned model trained by supervised learning to generate the reference operational data from the training sound data, the training sound data including training sound data recorded while a reference vehicle in a state serving as a reference for evaluation is operated for a predetermined time, and reference operational data indicating the operational status of the reference vehicle collected simultaneously with the training sound data. The processing circuit executes a generation process of inputting the evaluation sound data recorded while the target vehicle is operated for the predetermined time into the trained model and outputting generated data, which is data on the operational status, and an evaluation process of comparing the generated data with target operational data indicating the operational status of the target vehicle collected simultaneously with the evaluation sound data, and evaluating the target vehicle according to the magnitude of the difference between the target operational data and the generated data. Effect of the Invention

[0006] The difference between the condition of the target vehicle and the reference vehicle appears in the magnitude of the deviation between the generated data and the target operational data. Therefore, the vehicle evaluation system can determine the difference in the condition of the vehicle from the sound data and perform the evaluation. [Brief description of the drawings]

[0007] [Figure 1] FIG. 1 is a schematic diagram showing an embodiment of a vehicle evaluation system. [Diagram 2] FIG. 2 is a schematic diagram showing the configuration of the vehicle control unit. [Diagram 3] FIG. 3 is a flowchart of the data acquisition process. [Figure 4] FIG. 4 is a flowchart of the data formatting process. [Diagram 5] FIG. 5 is a flowchart of the training process. [Figure 6]FIG. 6 is a flowchart relating to the generation process and the evaluation process. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] Hereinafter, an embodiment of a vehicle evaluation system will be described with reference to FIGS. <Vehicle evaluation system configuration> As shown in Fig. 1, this vehicle evaluation system includes a data center 100 and a data acquisition device 300. The data center 100 is communicatively connected to the data acquisition device 300 via a communication network 200. As shown in Fig. 1, the data center 100 includes a storage device 120 in which a program is stored, and a processing circuit 110. The processing circuit 110 executes the program stored in the storage device 120 to perform various processes. The data center 100 also includes a communication device 130.

[0009] The data acquisition device 300 is, for example, a personal computer. The data acquisition device 300 includes a storage device 320 in which a program is stored, and a processing circuit 310. The processing circuit 310 executes the program stored in the storage device 320 to execute various processes. The data acquisition device 300 also includes a communication device 330. In this embodiment, the data acquisition device 300 is connected to the data center 100 by wireless communication via the communication network 200. The data acquisition device 300 also includes a display device 340 that displays information. The data acquisition device 300 also includes a microphone 350.

[0010] When evaluating a target vehicle 10 using this vehicle evaluation system, a microphone 350 is installed at a predetermined position relative to the target vehicle 10. In addition, a data acquisition device 300 is connected to the vehicle control unit 20 of the target vehicle 10. Then, an operator operates the target vehicle 10 to operate the target vehicle 10. While the target vehicle 10 is being operated in this manner, the data acquisition device 300 records sounds with the microphone 350. In addition, the data acquisition device 300 acquires target operation data indicating the operating status of the target vehicle 10 at the same time as recording the sound data.

[0011] As shown in FIG. 1, the vehicle control unit 20 includes a storage device 22 in which a program is stored, and a processing circuit 21. The processing circuit 21 executes various controls by executing the programs stored in the storage device 22. The vehicle control unit 20 controls each part of the target vehicle 10. The vehicle control unit 20 is connected to various sensors that detect the state of the target vehicle 10. A crank position sensor 34 is connected to the vehicle control unit 20. The crank position sensor 34 outputs a crank angle signal according to a change in the rotation phase of a crankshaft, which is an output shaft of an internal combustion engine mounted on the target vehicle 10. The vehicle control unit 20 calculates an engine rotation speed NE, which is the rotation speed of the crankshaft, based on the crank angle signal. An air flow meter 33 is connected to the vehicle control unit 20. The air flow meter 33 detects an intake air temperature THA, which is the temperature of air drawn into a cylinder through an intake passage of the internal combustion engine mounted on the target vehicle 10, and an intake air amount Ga, which is the mass of the drawn air.

[0012] The vehicle control unit 20 is also connected to a transmission control unit 30 that controls the transmission mounted on the target vehicle 10. The vehicle control unit 20 acquires information on the gear ratio, the input rotation speed Nin, the output rotation speed Nout, and the oil temperature of the transmission from the transmission control unit 30. The input rotation speed Nin is the rotation speed of the input shaft of the transmission. The output rotation speed Nout is the rotation speed of the output shaft of the transmission. When the data acquisition device 300 is connected to the vehicle control unit 20 of the target vehicle 10, the data acquisition device 300 becomes able to acquire information on the target vehicle 10 through the vehicle control unit 20.

[0013] <Evaluation process using the vehicle evaluation system> As described above, in this vehicle evaluation system, when evaluating the target vehicle 10, the data acquisition device 300 is connected to the vehicle control unit 20 of the target vehicle 10. Then, while the target vehicle 10 is operating, the data acquisition device 300 records sounds with the microphone 350. The data acquisition device 300 transmits data including the recorded sound data to the data center 100. Then, the data center 100 executes an evaluation process for evaluating the target vehicle 10 using the received data.

[0014] The data acquisition device 300 records the sound data collected by the microphone 350 while the target vehicle 10 is operated for a predetermined time in the storage device 320 as evaluation sound data. The data acquisition device 300 also stores the intake air temperature THA detected by the air flow meter 33 in the storage device 320 as information on the outside air temperature when the sound data is recorded. The data acquisition device 300 also stores the target operation data collected simultaneously with the sound data in the storage device 320. The target operation data is, for example, the engine speed NE, the input rotation speed Nin, the output rotation speed Nout, and the gear ratio. The data acquisition device 300 then stores the evaluation sound data, the outside air temperature data, and the target operation data collected in this manner in the storage device 320 as one data set for a predetermined time.

[0015] The data acquisition device 300 cuts out data from a data set for a predetermined time stored in the storage device 320 for each data range of a window Tw having a time width shorter than the predetermined time, and shapes the data into evaluation data. In the data shaping process for shaping the evaluation data, the data acquisition device 300 converts the evaluation sound data into a mel spectrogram obtained by frequency analysis and handles it as image data. In the mel spectrogram, the vertical axis is frequency, and the frequency on the vertical axis is shown in the mel scale. The horizontal axis is the time axis. In the mel spectrogram, intensity is represented by color. The lower the intensity, the darker the blue color, and the higher the intensity, the brighter the red color. The sound data included in one data set for the predetermined time becomes one mel spectrogram for the predetermined time. The data acquisition device 300 transmits the shaped evaluation data to the data center 100. Then, the data center 100 inputs the evaluation sound data included in the evaluation data shaped into a plurality of lists into a learned model trained by supervised learning, and performs evaluation processing.

[0016] <About trained models> The storage device 120 of the data center 100 stores data of a trained model used for evaluating the target vehicle 10. In the data center 100, in order to handle the evaluation sound data as image data, a model that partially uses ResNet-18, which is an image classification model, is used. Note that ResNet-18 is a pre-trained image classification model trained on the ImageNet dataset. ResNet-18 is trained on more than 1 million pieces of image data and can classify input images into 1000 categories. The trained model stored in the storage device 120 of the data center 100 is a model obtained by transfer learning the pre-trained ResNet-18. This trained model is a model obtained by replacing the classification output layer of ResNet-18 with a neural network MLP and training this neural network MLP by supervised learning. The input layer Lin of the neural network MLP includes a first input layer Lin1 to which the output from ResNet-18 is input, as well as a second input layer Lin2. As a result, the vehicle evaluation system can reflect data other than the evaluation sound data included in the evaluation data in the evaluation. For example, the second input layer Lin2 includes a node for inputting data on outside air temperature.

[0017] In this vehicle evaluation system, evaluation sound data acquired while operating the target vehicle 10 for a predetermined time is input to a trained model, and a generation process is executed to generate data on the operating status of the target vehicle 10 as generated data. The operating status data includes the engine rotation speed NE, the input rotation speed Nin, the output rotation speed Nout, and the gear ratio. The output layer Lout of the neural network MLP is composed of four nodes that output these values.

[0018] Next, a training process for training the model to obtain a trained model will be described. The model is trained by supervised learning using a large amount of measurement data collected in advance using a reference vehicle that serves as the basis for evaluation. In this example, the reference vehicle is a vehicle that has been subjected to a certain amount of break-in driving after manufacture, and has been thoroughly inspected and confirmed to have no abnormalities. In other words, the reference vehicle is a vehicle in extremely good condition with almost no deterioration.

[0019] <About data acquisition processing> The data acquisition process for acquiring the measurement data is executed by a computer that can be connected to the vehicle control unit 20 and acquire data, similar to the data acquisition device 300.

[0020] FIG. 3 is a flowchart showing the flow of the data acquisition process for acquiring measurement data. In the following, the step number of each process is represented by a number with "S" at the beginning. This data acquisition process is executed while the reference vehicle is operating. As shown in FIG. 3, when the data acquisition process is started, the computer starts measuring data (S100). Then, while operating the reference vehicle, the computer records sound data with the microphone 350 and acquires outside air temperature data and reference operating data. In addition, in order to train the model, this sound data that records the sound emitted by the reference vehicle is training sound data. Next, the computer determines whether or not the collection of measurement data over a predetermined time has been completed (S110). If the computer determines that the collection of measurement data has not been completed (step S110: NO), the computer repeats the process of S110. On the other hand, if the computer determines that the collection of measurement data has been completed (S110: YES), the computer proceeds to the process of S120. Then, the computer ends the measurement of data (S120). When the measurement is thus completed, the computer records the measurement data for a predetermined time period as one data set in the storage device (S130). Then, the computer temporarily ends this data acquisition process. Thus, acquisition of one data set is completed. Training data used in supervised learning is collected by repeating this data acquisition process many times to collect a large number of data sets of measurement data recorded while operating the reference vehicle. The data set of measurement data collected in this manner is shaped through a data shaping process shown in FIG. 4.

[0021] <About data formatting processing> The data shaping process is a process of cutting out one data set while shifting the window Tw and shaping the data into multiple lists. This data shaping process is performed by a computer. The computer that performs this data shaping process may be the same as the computer that executes the data acquisition process, or may be a different computer.

[0022] When the data shaping process starts, the computer first reads one data set of measurement data (S200). Next, the computer converts the sound data in the data set read in the process of S200 into a mel spectrogram (S210). Next, the computer standardizes data other than the sound data in the data set (S220).

[0023] Next, the computer sets the data extraction start time t to 0 (S230). Then, the computer extracts the data (S240). That is, the computer aligns the start point of the window Tw with the extraction start time t, and extracts the data within the window Tw. Specifically, the computer extracts an image within the window Tw range from the mel spectrogram. The computer extracts data within the window Tw range from the outside air temperature data and the operating status data.

[0024] Next, the computer calculates a representative value of each data extracted through the process of S240 (S250). For example, the average value of the data in the window Tw is calculated as the representative value in this window Tw. Instead of the average value, the maximum value or the minimum value may be calculated as the representative value. Next, the computer determines whether or not to shift the window Tw by the stride t_st (S260). Data extraction is performed by repeatedly shifting the window Tw by the stride t_st for the data set acquired during the measurement pattern. When the window Tw reaches the end of the data set and all the data contained in the data set has been extracted, the window Tw can no longer be shifted by the stride t_st. When the computer is unable to shift the window Tw by the stride t_st in this way, it makes a negative determination in the process of S260.

[0025] When the computer determines that the window Tw is shifted by the stride t_st (S260: YES), the computer stores the data of the cut-out image and a set of each representative value in one list (S270). When storing in the list, the computer resizes the image data to a size of 224×224 suitable for input to ResNet-18. Then, the computer updates the cut-out start time t (S280). Specifically, the cut-out start time t is updated by adding the stride t_st to the cut-out start time t to set the sum as the new cut-out start time t. As a result, the window Tw is shifted by the stride t_st. Then, the computer shifts the window Tw by the stride t_st and executes the processes of S240 to S260 again. That is, the computer repeats the processes of S240 to S280 until the window Tw cannot be shifted by the stride t_st.

[0026] When the computer determines that the window Tw cannot be shifted (S260: NO), the computer advances the process to S290. The process of S290 is the same as the process of S270. In the process of S290, when the data is stored in the list, the computer determines whether all the loaded data sets have been processed (S300). When the computer determines that the processing of all the data sets has not been completed (S300: NO), the computer returns the process to S200. Then, the computer reads one unprocessed data set and executes the processes from S210 onward. When the computer determines that all the data sets have been processed (S300: YES), the computer ends this series of data shaping processes. In this way, the computer shapes the prepared data sets of measurement data into the form of a plurality of lists. In this way, a training process is performed to train a model using a large number of data sets that have been shaped into a set of a plurality of lists through the data shaping process.

[0027] <About the training process> FIG. 5 is a flowchart of the training process. This training process is performed by a computer. A large amount of data sets shaped through a data shaping process are stored as training data in a storage device of the computer that executes the training process. When the training process is started, the computer first reads the training data stored in the storage device (S400). Next, the computer reads one data set from the read training data (S410). Next, the computer reads one list from the data set (S420).

[0028] Then, the computer inputs the image data and the outside air temperature data from the list data into the above-mentioned model to calculate the data of the operating status (S430). When the training process is started, the ResNet-18 part has already been trained, but the weights and biases of the neural network MLP part are set to initial values. In this training process, the weights and biases of the neural network MLP part are updated. Specifically, the image data Dw resized to a size of 224×224 included in the list is input into ResNet-18. Then, a representative value of the outside air temperature data is input into the second input layer Lin2 of the neural network MLP. The feature amount of the image data Dw is extracted through ResNet-18 and input into the first input layer Lin1 of the neural network MLP. Then, the value of the data of the operating status is output from the output layer Lout of the neural network MLP. When the data of the operating status is calculated in this way, the computer records the value of the data of the operating status (S440).

[0029] Next, the computer determines whether all the lists included in the data set have been processed (S450). If it determines that processing of all the lists has not been completed (S450: NO), the computer returns the process to S420. Then, the computer reads one unprocessed list (S420) and executes the process from S430 onward.

[0030] If it is determined that all the lists have been processed (S450: YES), the computer proceeds to S460. In this manner, the computer calculates the value of the operational status data for each of the lists included in the loaded data set (S430). Then, the calculated values ​​are recorded (S440).

[0031] In the process of S460, the computer calculates an evaluation index value. The evaluation index value is a value indicating the magnitude of deviation between the value calculated through the process of S430 and the data of the operating status included in the dataset. The data of the operating status included in the dataset is reference operating data and is a correct value. Here, the computer calculates the magnitude of deviation between the value calculated through the process of S430 and the correct value. After calculating the evaluation index value in this way, the computer performs learning (S470). Specifically, the computer adjusts the weights and biases in the neural network MLP by the backpropagation method so that the value of the evaluation index value becomes smaller. Next, the computer determines whether all the datasets included in the loaded training data have been processed (S480). If it is determined that the processing of all the datasets has not been completed (S480: NO), the computer returns the process to S410. Then, the computer reads one unprocessed dataset (S410) and executes the processes from S420 onwards. The computer repeats learning in this way to train the model until the processing of all the datasets is completed. In this way, the computer trains a model through supervised learning so as to be able to generate reference operational data, which is data indicating operational conditions, from image data and outside temperature data.

[0032] When it is determined that all the data sets have been processed (S480: YES), the computer records the parameters of the model that has completed learning using all the data sets in the storage device (S490). Then, the computer ends this series of training processes. In this way, data of the trained model is obtained through the training process. The data of the trained model trained in this way through the training process is stored in the storage device 120 of the data center 100.

[0033] Next, the data acquisition process, data formatting process, and evaluation process when evaluating the target vehicle 10 using the vehicle evaluation system will be described. <Data Acquisition Process by Data Acquisition Device 300> When evaluating a vehicle using the vehicle evaluation system, as described above, the data acquisition device 300 is connected to the target vehicle 10, which is the vehicle to be evaluated. In addition, the microphone 350 is installed in the target vehicle 10. Then, an operator operates the target vehicle 10 to operate the target vehicle 10. While the target vehicle 10 is thus operated, the data acquisition device 300 records sound with the microphone 350. In addition, the data acquisition device 300 simultaneously acquires outside air temperature data. In addition, the data acquisition device 300 acquires target operation data indicating the operating status of the target vehicle 10 at the same time as the recording of the sound data. Specifically, the data acquisition device 300 executes the data acquisition process described with reference to FIG. 3 to acquire one data set as evaluation data. The sound data included in the evaluation data is evaluation sound data.

[0034] <Data shaping process by data acquisition device 300> The data acquisition device 300 executes a data shaping process to shape the evaluation data. Specifically, the data shaping process described with reference to FIG. 4 is executed to shape the evaluation data into a plurality of lists. Since only one evaluation data is acquired in the data acquisition process, only one data set is read and shaped in the data shaping process at this time. Through this data shaping process, the evaluation data is shaped into a plurality of lists of image data Dw, outside air temperature data, and target operation data. After executing the data shaping process to shape the evaluation data, the data acquisition device 300 transmits the shaped evaluation data to the data center 100.

[0035] <Evaluation process by data center 100> Upon receiving the evaluation data, the data center 100 stores the evaluation data in the storage device 120. Then, the data center 100 executes an evaluation process for evaluating the target vehicle 10 by executing a routine shown in FIG. 6. The routine shown in FIG. 6 is executed by the processing circuit 110 of the data center 100. As shown in FIG. 6, when this routine is started, the data center 100 reads the evaluation data stored in the storage device 120 (S500). Then, through the processes of S510 to S540, the data center 100 repeats the calculation of the data of the operating status using the trained model stored in the storage device 120. Specifically, the data center 100 reads one list from the evaluation data (S510). The data center 100 inputs the data of the list into the trained model to calculate the data of the operating status (S520), similar to the process of S430 in the training process. When the data of the operating status is calculated in this manner, the data center 100 records the value of the data of the operating status (S530).

[0036] Next, the data center 100 determines whether all the lists included in the data set have been processed (S540). If it is determined that processing of all the lists has not been completed (S540: NO), the data center 100 returns the process to S510. Then, the data center 100 reads one unprocessed list (S510) and executes the processes from step S520 onwards. If it is determined that all the lists have been processed (S540: YES), the data center 100 advances the process to S550.

[0037] In this way, the data center 100 calculates the value of the data of the operating status for each of the lists included in the read evaluation data (S520). Then, the calculated value is recorded (S530). This series of processes from S510 to S540 is a generation process in which the evaluation sound data is input to the trained model and generated data is output. Next, the data center 100 calculates an evaluation index value in the same manner as the process of S460 in the training process (S550). The trained model is optimized to generate generated data in which the reference operating data is restored from the sound emitted from the reference vehicle. Therefore, when evaluation sound data emitted from the target vehicle 10 in a state different from the reference vehicle is input, the data of the operating status cannot be restored correctly. In other words, if the state of the target vehicle 10 deviates from the state of the reference vehicle, a deviation occurs between the target operating data stored in the data set as the correct data and the generated data. The evaluation index value indicates the magnitude of this deviation. In other words, the larger the evaluation index value, the more the state of the target vehicle 10 deviates from the state of the reference vehicle. As described above, the reference vehicle is a vehicle in extremely good condition with almost no deterioration. Therefore, in this evaluation system, the smaller the evaluation index value, the closer the condition of the target vehicle 10 is to the condition of the reference vehicle, and the higher the evaluation. After calculating the evaluation index value in this way, the data center 100 proceeds to S560.

[0038] The data center 100 determines the evaluation rank based on the evaluation index value (S560). The data center 100 determines the evaluation rank by selecting an evaluation rank according to the magnitude of the evaluation index value from among four evaluation ranks, S rank, A rank, B rank, and C rank. The S rank is an evaluation rank indicating that the vehicle is the highest rated vehicle among the four evaluation ranks. The C rank is an evaluation rank indicating that the vehicle is the lowest rated vehicle among the four evaluation ranks. The evaluations are S rank, A rank, B rank, and C rank in descending order, with C rank being the lowest rated evaluation rank. After determining the evaluation rank of the target vehicle 10 through the processing of S560, the data center 100 advances the processing to S570. The processing of S550 and the processing of S560 are evaluation processing for comparing the target operating data with the generated data and evaluating the target vehicle 10 according to the magnitude of the deviation between the target operating data and the generated data.

[0039] Next, the data center 100 transmits the evaluation rank to the data acquisition device 300 and outputs the evaluation rank (S570). After outputting the evaluation rank in this manner, the data center 100 ends this routine.

[0040] The data acquisition device 300 that receives the evaluation rank displays the received evaluation rank on the display device 340 as the evaluation rank of the target vehicle 10 . <Action of this embodiment> The data center 100 executes a generation process using the trained model to generate generated data by restoring operation data from the evaluation data. The trained model is a neural network in which the feature quantities of data extracted from a predetermined time worth of data are used as explanatory variables, and the operating status at the time point corresponding to the extracted data is used as a target variable. The data center 100 then executes an evaluation process based on the generated data. The reference operating data and the target operating data are the engine rotation speed NE, the input rotation speed Nin, the output rotation speed Nout, and the gear ratio. That is, the reference operating data and the target operating data include data on the rotation speed of a rotating shaft in the powertrain.

[0041] In the evaluation process, the data center 100 calculates, as an evaluation index value, the sum of the deviations output for a predetermined time period for each piece of data repeatedly extracted while changing the extraction start time t. The data center 100 determines the evaluation rank of the target vehicle 10 based on the evaluation index value.

[0042] <Effects of this embodiment> (1) The difference between the condition of the target vehicle 10 and the reference vehicle appears in the magnitude of the deviation between the generated data and the target operational data. Therefore, the vehicle evaluation system can determine the difference in the condition of the vehicle from the sound data and perform the evaluation.

[0043] (2) The vehicle evaluation system uses image data obtained by frequency analysis of sound data, so that the vehicle evaluation system can efficiently extract features contained in the sound data and perform evaluation processing.

[0044] (3) The above vehicle evaluation system divides data for a given time into multiple sections and analyzes them. The results are then integrated to calculate the evaluation index value. Therefore, according to the above vehicle evaluation system, the trained model can be smaller than when analyzing data for a given time all at once.

[0045] (4) The vehicle evaluation system outputs the evaluation results by applying them to a preset evaluation rank. Therefore, this evaluation system makes it easier to understand the relative level of the condition of the target vehicle 10 in the used car market.

[0046] (5) Sound data may be affected by differences in the measurement environment even if the vehicle condition is the same. The above vehicle evaluation system includes outside temperature data in the training data and evaluation data. Therefore, the vehicle evaluation system can perform evaluations that reflect the effects of differences in outside temperature.

[0047] (6) A method of comparing sound data measured in a reference vehicle with sound data measured in a target vehicle 10 and calculating the degree of deviation between these sound data as an evaluation index value is also conceivable. However, in the case of such a method, if the measurement conditions vary due to the variation in the operation by the operator, the sound data will vary. Therefore, in the case of such a method, the influence of the variation in the measurement conditions is reflected in the evaluation index value. In contrast, the vehicle evaluation system of the above embodiment uses evaluation data collected by operating the target vehicle 10. Then, this vehicle evaluation system compares the target operation data included in the evaluation data with the generated data generated using the evaluation sound data. Therefore, the influence of the variation in the measurement conditions can be suppressed.

[0048] <Example of change> This embodiment can be modified as follows: This embodiment and the following modifications can be combined with each other to the extent that there is no technical contradiction.

[0049] The data shaping process may be performed in the data center 100. The data used as the measurement data and evaluation data does not have to be mel spectrograms. For example, a spectrogram obtained by wavelet transforming sound data may be used. A spectrogram obtained by short-time Fourier transforming sound data may also be used. It is not necessary to convert sound data into image data. For example, features may be extracted from the sound data itself and used as the measurement data and evaluation data. In that case, the model used for the evaluation process does not need to use ResNet-18, which handles image data. In addition, although a model in which ResNet-18 is transfer-trained has been exemplified, the structure of the model is not limited to this structure. It is sufficient for the trained model to be able to output generated data based on the evaluation data.

[0050] The number of evaluation ranks does not have to be four. For example, the number of evaluation ranks may be increased. Conversely, the number of evaluation ranks may be decreased. As the evaluation process, an example in which the evaluation rank is determined based on the evaluation index value has been shown, but the evaluation process is not limited to this mode. For example, the evaluation index value may be output as it is, with the larger the value, the lower the evaluation, and displayed on the display device 340.

[0051] In the above vehicle evaluation system, an example has been given in which a vehicle in extremely good condition is used as the reference vehicle to evaluate the target vehicle 10. It is not necessary to use a vehicle in good condition as the reference vehicle. For example, a vehicle in extremely poor condition with a low rating can also be used as the reference vehicle. The evaluation index value in the above evaluation process is a value indicating the degree of deviation between the condition of the target vehicle 10 and that of the reference vehicle. Therefore, when an aged vehicle with an extremely low rating is used as the reference vehicle, the smaller the evaluation index value, the lower the rating. The target vehicle 10 can also be evaluated using such evaluation index values.

[0052] The vehicle evaluation system may be configured only by the data center 100 that performs the generation process and the evaluation process. In this case, the vehicle evaluation system is a system that performs the generation process and the evaluation process using the received evaluation data and outputs the evaluation result. Also, for example, the data of the trained model may be stored in the storage device 320 of the data acquisition device 300, and the vehicle evaluation system may be configured only by the data acquisition device 300. In this case, the generation process and the evaluation process are executed in the data acquisition device 300.

[0053] The data acquisition device 300 itself does not have to include the microphone 350. Sound data can be acquired from an external device and data shaping, generation, and evaluation can be performed. The generation and evaluation can be performed using multiple sound data recorded using multiple microphones 350.

[0054] In the above embodiment, an example has been described in which the vehicle evaluation system evaluates the target vehicle 10 itself. In contrast, the vehicle evaluation system may evaluate the target vehicle 10 by evaluating a specific unit in the target vehicle 10. For example, the vehicle evaluation system may evaluate the transmission mounted on the target vehicle 10 using sound data recorded from sounds emitted from the transmission. [Explanation of symbols]

[0055] 10... target vehicle, 20... vehicle control unit, 21... processing circuit, 22... storage device, 30... transmission control unit, 100... data center, 110... processing circuit, 120... storage device, 130... communication device, 200... communication network, 300... data acquisition device, 310... processing circuit, 320... storage device, 330... communication device, 340... display device, 350... microphone

Claims

1. A vehicle evaluation system for evaluating a target vehicle using sound data recorded from a target vehicle, the target vehicle being an evaluation target vehicle, the vehicle evaluation system comprising: a processing circuit; and a storage device; The storage device stores data of a trained model trained by supervised learning to generate the reference operational data from the training sound data using training data including training sound data recorded while a reference vehicle in a state serving as a reference for evaluation is operated for a predetermined time period and reference operational data indicating the operational status of the reference vehicle collected simultaneously with the training sound data; The processing circuit executes a generation process of inputting evaluation sound data recorded while operating the target vehicle for the predetermined time into the trained model and outputting generated data, which is data on the operating status, and an evaluation process of comparing target operating data indicating the operating status of the target vehicle, which was collected simultaneously with the evaluation sound data, with the generated data, and evaluating the target vehicle according to the magnitude of deviation between the target operating data and the generated data. Vehicle rating system.

2. The trained model generates the reference operational data from evaluation sound data including image data of a spectrogram obtained by frequency analysis of the sound data. The vehicle evaluation system according to claim 1 .

3. The trained model is a neural network that uses features of data extracted from the data for the predetermined time period as explanatory variables and uses the operation status at the time point corresponding to the extracted data as a target variable. The vehicle evaluation system according to claim 1 .

4. The evaluation process calculates, as an evaluation index value, a sum of the deviations output for the predetermined time period for each piece of data repeatedly extracted while changing the extraction start time. The vehicle evaluation system according to claim 3 .

5. The reference operation data and the target operation data include data on the rotation speed of a rotating shaft in a power train. The vehicle evaluation system according to claim 1 .

Images