JP7320755B2 - Vehicle simulation system, vehicle simulation method and computer program - Google Patents

Description

TECHNICAL FIELD The present disclosure relates to data processing technology, and more particularly to a vehicle simulation system, vehicle simulation method and computer program.

In the development and evaluation phases of in-vehicle equipment, a vehicle simulation system that simulates the behavior of an actual vehicle is sometimes used instead of using an actual vehicle in order to reduce the costs required for development and evaluation.

Japanese translation of PCT publication No. 2018-514042

In previous vehicle simulation systems, only the vehicle position (which can be said to be the ideal position) generated by the vehicle model was depicted in the simulation results. Therefore, the simulation results may not contribute sufficiently to the evaluation of the actual system.

The present disclosure has been made in view of such circumstances, and one object thereof is to enhance the usefulness of simulation results in a vehicle simulation system.

In order to solve the above-described problems, a vehicle simulation system according to one aspect of the present disclosure uses, as an explanatory variable, information indicating vehicle behavior that is input to a vehicle when the vehicle is actually traveling, and the behavior of the vehicle indicated by the information; A storage unit that stores a model whose objective variable is the difference between the vehicle's actual behavior and ideal values for the vehicle's behavior derived from the simulation scenario. A deriving unit for deriving an estimated value regarding the actual behavior of the vehicle in a situation, and a generating unit for generating an image showing the ideal value and the estimated value in different modes as an image showing the behavior of the vehicle in the simulation.

Another aspect of this disclosure is also a vehicle simulation system. This vehicle simulation system uses information that indicates the behavior of the vehicle input to the vehicle when the vehicle is actually running as an explanatory variable, and uses the difference between the behavior of the vehicle indicated by the information and the actual behavior of the vehicle as the objective variable. A storage unit that stores a model and values corresponding to ideal values regarding vehicle behavior derived based on a simulation scenario. A first derivation unit that derives a quasi-ideal value, which is a value to be transmitted through the model, and an estimated value regarding the actual behavior of the vehicle in the situation indicated by the simulation scenario is derived by inputting the ideal values regarding the behavior of the vehicle into the model. and a generator that generates an image showing the quasi-ideal value and the estimated value in different modes as an image showing the behavior of the vehicle in the simulation.

Yet another aspect of the present disclosure is a vehicle simulation method. This method uses a model stored in a storage unit, which is input to the vehicle when the vehicle is actually running, as an explanatory variable to indicate the behavior of the vehicle. By inputting the ideal values of vehicle behavior derived based on the simulation scenario into a model whose objective variable is the difference from the behavior of the vehicle, the estimated values of the actual behavior of the vehicle under the conditions indicated by the simulation scenario can be obtained. A computer performs the derivation and generates images showing the ideal and estimated values in different ways as images showing the behavior of the vehicle in the simulation.

Yet another aspect of the present disclosure is also a vehicle simulation method. In this method, values corresponding to ideal values regarding vehicle behavior derived based on a simulation scenario are transmitted via an in-vehicle network to cause the vehicle to execute the behavior indicated by the ideal values. A quasi-ideal value that is a value is derived, and is a model stored in a storage unit, and is an explanatory variable that indicates the behavior of the vehicle that is input to the vehicle when the vehicle is actually traveling. By inputting the ideal values of vehicle behavior into a model whose objective variable is the difference between the actual behavior of the vehicle and the actual behavior of the vehicle, an estimated value of the actual behavior of the vehicle in the situation indicated by the simulation scenario is derived, and the simulation The computer generates an image showing the quasi-ideal value and the estimated value in different manners as an image showing the behavior of the vehicle in .

Any combination of the above components and expressions of the present disclosure converted between devices, computer programs, recording media recording computer programs, vehicles equipped with vehicle support systems, etc. are also included in the present disclosure. It is effective as an aspect.

According to the present disclosure, it is possible to enhance the usefulness of simulation results in a vehicle simulation system.

1 is a block diagram showing functional blocks of a conventional vehicle simulation system; FIG. FIG. 4 is a diagram showing an example of sample data for constructing a sensor model in the first embodiment; FIG. 2 is a block diagram showing functional blocks of the sensor model generation device of the first embodiment; FIG. 1 is a block diagram showing functional blocks of a vehicle simulation system of a first embodiment; FIG. 4 is a block diagram showing details of a sensor model unit of the first embodiment; FIG. FIG. 5 is a diagram showing an example of curve fitting; It is a figure which shows the state of the vehicle periphery in a modification. FIG. 10 is a diagram showing an example of sample data for constructing a sensor model in the second embodiment; FIG. 11 is a block diagram showing functional blocks of a sensor model generation device of a second embodiment; FIG. 11 is a block diagram showing details of a sensor model unit of the second embodiment; FIG. 11 is a block diagram showing functional blocks of a sensor model generation device of a third embodiment; FIG. 11 is a block diagram showing details of a sensor model unit of the third embodiment; FIG. 11 is a block diagram showing functional blocks of a sensor model generation device of a fourth embodiment; FIG. 11 is a block diagram showing functional blocks of a vehicle simulation system of a fourth embodiment; FIG. FIG. 11 is a block diagram showing functional blocks of a simulation control unit of a fourth embodiment; FIG. 13 is a block diagram showing details of a sensor model unit of the fourth embodiment; FIG. 10 is a diagram showing an example of a simulation result image; FIG. 10 is a diagram showing an example of a simulation result image; FIG. 10 is a diagram showing an example of a simulation result image; FIG. 10 is a diagram showing an example of a simulation result image;

Give an overview. In the first to third embodiments, a sensor model generating device and a vehicle simulation system using a model generated by the sensor model generating device are proposed as techniques for supporting vehicle simulation.

The sensor model generation devices of the first to third embodiments include a first derivation section, a second derivation section, and a generation section. The first derivation unit derives an attribute state related to automatic driving of the vehicle based on the output result of the sensor mounted on the vehicle. The second derivation unit derives the state of attributes related to the vehicle with higher accuracy than the first derivation unit. The generation unit is a model for simulating a vehicle, and the state derived by the first derivation unit or the second derivation unit is used as an explanatory variable, and the state derived by the second derivation unit and the state derived by the first derivation unit Generate a model with the difference from the derived state as the objective variable.

Further, the vehicle simulation system of the first to third embodiments includes a storage section, an estimation section, and a simulation section. The storage unit stores the state of the attribute derived by the derivation unit, or the state of the attribute, with respect to the derivation unit that derives the state of the attribute related to automatic driving of the vehicle based on the output result of the sensor mounted on the vehicle. A model for estimating the true value of the attribute state derived by the derivation unit or the true value of the attribute state as an explanatory variable, and the true value of the attribute state and the true value of the attribute state by the derivation unit A model is stored in which the objective variable is the difference from the derived state of the attribute. The estimating unit inputs one of the state of the attribute derived by the deriving unit and the true value of the state of the attribute into the model as a parameter of the simulation, thereby obtaining the state of the attribute derived by the deriving unit and the attribute Estimate the other of the true values of the states of The simulation section executes a vehicle-related simulation based on the estimation result of the estimation section.

<First embodiment>
FIG. 1 is a block diagram showing functional blocks of a conventional vehicle simulation system 100. As shown in FIG. Each block shown in the block diagram of the present disclosure can be realized by hardware such as a CPU and memory of a computer or a mechanical device, and is realized by a computer program or the like in terms of software. , and the functional blocks realized by their cooperation are drawn. It should be understood by those skilled in the art that these functional blocks can be implemented in various ways by combining hardware and software.

The vehicle simulation system 100 includes a simulation control section 10 , a user interface 12 , an environment data generation section 14 , an image analysis section 16 and a vehicle model section 18 . The simulation control unit 10 controls simulation of vehicle behavior. The simulation control unit 10 also transmits and receives data to and from the user interface 12 , environment data generation unit 14 , image analysis unit 16 and vehicle model unit 18 .

The user interface 12 provides an interface with the user (or the user's terminal). The environmental data generation unit 14 generates image data (CG) for analysis by the image analysis unit 16, which will be described later, according to parameters specified by the user (for example, the distance and direction to the pedestrian). For example, the environment data generator 14 generates CG indicating that there is a pedestrian 6 meters ahead of the vehicle.

The image analysis unit 16 analyzes the image data to detect the state of the vehicle surroundings (positions of obstacles, etc.) for controlling automatic driving of the vehicle. For example, the image analysis unit 16 analyzes a CG indicating that there is a pedestrian 6 meters ahead of the vehicle and detects that the distance between the vehicle and the pedestrian is 6.2 meters. In this case, there is an error of 0.2 meters. In the actual vehicle, the object of analysis by the image analysis unit 16 is the image captured by the vehicle-mounted camera, but in the vehicle simulation system 100, it is the CG generated by the environment data generation unit 14 .

The vehicle model unit 18 determines the behavior of the vehicle based on the analysis result of the image analysis unit 16, that is, simulates automatic driving of the vehicle. For example, if the image analysis unit 16 detects that there is a pedestrian 6.2 meters ahead of the vehicle in the direction in which the vehicle is traveling, the vehicle model unit 18 activates the brakes. You may output that the vehicle will stop in front of you. The simulation control unit 10 causes the simulation result by the vehicle model unit 18 to be displayed on a predetermined display device or stored in a predetermined storage device.

As described above, the conventional vehicle simulation system 100 creates an image showing the state of the surroundings of the vehicle, detects the state of the surroundings of the vehicle from the image, and simulates the details of automatic driving. In order to increase the comprehensiveness of the simulation, it is necessary to create many images corresponding to various cases such as brightness (weather, etc.) around the vehicle, road surface conditions, distance to obstacles, vehicle speed, etc. It takes a lot of time and money to create many images.

In addition, it is conceivable to create an image showing the state of the vehicle surroundings based on a theoretical model (for example, a theoretical calculation formula based on multiple parameters such as vehicle speed and brightness). The image may deviate from the real environment. As a result, the vehicle simulation results may deviate from the actual behavior of the vehicle.

Therefore, in the first embodiment, a mathematical model (hereinafter referred to as "sensor (also called "model"). The sensor model is then used to simulate vehicle behavior. This eliminates the need to create an image for each case in the simulation for detecting the state around the vehicle, reducing the cost of the simulation, and simulating realistic vehicle behavior.

First, the sensor model of the first embodiment will be explained. The sensor model of the first embodiment is a model for estimating a detection result based on an image (that is, a detection result by the image analysis unit 16 and an image analysis unit 28 to be described later) in a vehicle simulation system. FIG. 2 shows an example of sample data for constructing a sensor model in the first embodiment. The figure shows sample data for constructing a sensor model for pedestrian identification. Each sample contains the distance from the vehicle to the pedestrian and the weather as two factors that affect pedestrian identification accuracy.

The factor "distance" includes 10 categories corresponding to multiple steps of distance (true value) from the vehicle to the pedestrian. For example, one category may range from 0.1 meters to 2 meters, and another category may range from 2 meters to 4 meters. The factor "weather" includes three categories: sunny, cloudy, and rainy. For each sample, the value of the applicable category is "1" and the value of the non-applicable category is "0". This makes it possible to treat not only quantitative data but also qualitative data (in other words, quality-related data) as quantities. Factors are also called explanatory variables. Each category of factor is also called a dummy variable.

In addition, as data indicating the accuracy of image analysis, each sample is the difference between the true value of the distance from the vehicle to the pedestrian and the distance from the vehicle to the pedestrian detected by image analysis (also known as "detection distance error"). call). The sensor model of the example includes a plurality of dummy variables corresponding to a plurality of steps of distances from the vehicle to the detection target (pedestrian, etc.). Specifically, a total of 13 items, 10 categories of distance and 3 categories of weather, are used as dummy variables, and multiple regression analysis is performed with the detection distance error as the objective variable, to determine the effect of each category of each factor on the detection distance error. Generate a sensor model that indicates the magnitude of

式１のε_ｋは残差であり、ａ^ｉｊはｉ番目の要因のｊ番目のカテゴリを表すダミー変数にかかる係数である（以下「カテゴリ係数」とも呼ぶ。）。実施例のセンサモデル生成装置は、重回帰分析によって、残差の平方和が最小になるように各ダミー変数にかかるカテゴリ係数を導出する。 When the number _of factors is m, the number of categories of each factor is n ₁ , .

ε _k in Equation 1 is the residual, and a ^ij is the coefficient on the dummy variable representing the jth category of the ith factor (hereinafter also referred to as “category coefficient”). The sensor model generation device of the embodiment derives category coefficients for each dummy variable by multiple regression analysis so that the sum of squares of residuals is minimized.

FIG. 3 is a block diagram showing functional blocks of the sensor model generation device 20 of the first embodiment. The sensor model generation device 20 of the first embodiment includes a sensor data storage unit 22, a reference data storage unit 24, a tag data storage unit 26, an image analysis unit 28, a true value derivation unit 30, an error derivation unit 32, a true value quantization It has a unit 34 , a tag quantization unit 36 and a model generation unit 38 .

A computer program including a plurality of modules implementing the functions of the plurality of functional blocks shown in FIG. 3 may be stored in the storage of the sensor model generation device 20 . The CPU of the sensor model generation device 20 may read the computer program into the main memory and execute it, thereby exerting the functions of the functional blocks shown in FIG. Also, the plurality of functions shown in FIG. 3 may be distributed to a plurality of devices, and may be realized by the plurality of devices cooperating as a system. The same applies to other embodiments.

The sensor data storage unit 22 and the reference data storage unit 24 store data collected in tests using actual vehicles. The sensor data storage unit 22 stores the results of detection by various in-vehicle sensors for objects to be detected. In the embodiment, the sensor data storage unit 22 stores a plurality of image data captured by an in-vehicle camera. Image data showing the state, etc. are stored.

The reference data storage unit 24 stores a true value regarding the object of detection. In the embodiment, the reference data storage unit 24 stores data indicating the true value of the distance between the vehicle and the object, which is the result of detection by LIDAR (Light Detection and Ranging). The tag data storage unit 26 stores data set by users (developers, testers, etc.) as tag data. For example, the tag data storage unit 26 stores tag data indicating the weather on the day when the test using the vehicle was performed. The tag data may include various data such as road surface type (eg, asphalt or dirt), temperature, season, and the like.

The data stored in the sensor model generation device 20, the sensor data storage unit 22, and the reference data storage unit 24, which are collected or input in the same test, are associated with each other to form a set of sample data. . A plurality of sets of sample data (for example, 10,000 sets of sample data as shown in FIG. 2) are stored in the sensor model generation device 20, the sensor data storage unit 22, and the reference data storage unit 24. FIG.

The image analysis unit 28 corresponds to the image analysis unit 16 in FIG. 1 and also to the first derivation unit described in the outline above. That is, the image analysis unit 28 detects the state of the surroundings of the vehicle based on the image data stored in the sensor data storage unit 22 in order to control the automatic running of the vehicle. Specifically, the image analysis unit 28 detects the distance between the vehicle and the pedestrian in a plurality of samples based on the plurality of samples of image data stored in the sensor data storage unit 22 . A known technique may be adopted as a method of deriving the distance between the vehicle and the pedestrian based on the image.

The true value derivation unit 30 corresponds to the second derivation unit described in the outline above. The true value derivation unit 30 detects the state around the vehicle based on the LIDAR detection results stored in the reference data storage unit 24 . Specifically, the true value derivation unit 30 detects the distance between the vehicle and the pedestrian in a plurality of samples based on the LIDAR detection results of the plurality of samples stored in the reference data storage unit 24 . Here, distance detection using LIDAR is more accurate than distance detection using images, and substantially, the result of distance detection using LIDAR indicates the true value of the distance between the vehicle and the pedestrian. That is, it can be said that the true value derivation unit 30 detects the true value of the distance between the vehicle and the pedestrian.

For each sample, the error derivation unit 32 combines the detection result of the true value derivation unit 30 (the true value of the distance between the vehicle and the pedestrian in the embodiment) and the detection result of the image analysis unit 28 (the vehicle and the pedestrian in the embodiment). The detection distance error, which is the difference between the detection distance from the

For each sample, the true value quantization unit 34 converts the detection result of the true value derivation unit 30 (in the embodiment, the true value of the distance between the vehicle and the pedestrian) into one of a plurality of categories of predetermined explanatory variables. classified into For example, as shown in FIG. 2, the true value quantization unit 34 classifies the true value of the distance between the vehicle and the pedestrian detected by the true value derivation unit 30 into the corresponding category out of ten categories. You may

The tag quantization unit 36 classifies the tag data stored in the tag data storage unit 26 into one of a plurality of categories of predetermined explanatory variables for each sample. For example, the tag quantization unit 36 may classify the tag data indicating the weather into the corresponding category among the three categories, as shown in FIG.

The model generator 38 corresponds to the generator described in the outline above. The model generation unit 38 creates a combination of the error calculated by the error derivation unit 32, the category selected by the true value quantization unit 34, and the category selected by the tag quantization unit 36 for each sample. The model generation unit 38 uses the error calculated by the error derivation unit 32 as an objective variable, and statistically processes the category selected by the true value quantization unit 34 and the category selected by the tag quantization unit 36 as dummy variables. Multiple category coefficients for multiple categories are derived by performing (multiple regression analysis in the example). The model generating unit 38 generates the regression equation (equation 1 above) in which the category coefficients are set as a sensor model, and stores data of the sensor model in the model storage unit 40 .

Next, a vehicle simulation system using the sensor model generated by the sensor model generation device 20 will be described. FIG. 4 is a block diagram showing functional blocks of the vehicle simulation system 110 of the first embodiment. A vehicle simulation system 110 of the embodiment includes a simulation control section 10 , a user interface 12 , an environment data generation section 14 , a vehicle model section 18 and a sensor model section 50 .

A computer program including modules implementing the functions of the functional blocks shown in FIG. 4 may be stored in the storage of vehicle simulation system 110 . The CPU of vehicle simulation system 110 (or the CPU of a device within the system) may read out this computer program into the main memory and execute it, thereby exerting the function of each functional block shown in FIG. Also, the plurality of functions shown in FIG. 4 may be distributed to a plurality of devices, and may be realized by the plurality of devices cooperating as a system. Furthermore, multiple functions shown in FIG. 4 may be aggregated into a single device. The same applies to other embodiments.

Among the functional blocks of the vehicle simulation system 110 of the first embodiment, the functional blocks that are the same as or correspond to the functional blocks of the conventional vehicle simulation system 100 shown in FIG. 1 are denoted by the same reference numerals. In the following, re-description of the contents already described with reference to FIG. 1 will be omitted as appropriate.

The environment data generation unit 14 generates environment data, which is a premise of simulation, according to parameters received by the user interface 12 or parameters input from an external source such as a file. The environmental data includes the values of the explanatory variables of the sensor model. The environment data of the first embodiment includes the distance from the vehicle to the pedestrian as true value data, and the value indicating the weather as tag data.

The sensor model unit 50 corresponds to the estimation unit described in the outline above, and provides a function that replaces the image analysis unit 16 in the conventional vehicle simulation system 100 . The sensor model unit 50 inputs the true value of the state around the vehicle (the distance from the vehicle to the pedestrian in the embodiment) as a parameter for vehicle simulation into the sensor model, so that the image analysis unit 16 generates an image showing the above state. It functions as an estimation unit that estimates the detection result when analyzing

The vehicle model section 18 corresponds to the simulation section described in the outline above. The vehicle model unit 18 determines the behavior of the vehicle based on the estimation results from the sensor model unit 50, that is, simulates automatic driving of the vehicle. For example, when the sensor model unit 50 finds that there is a pedestrian 6.2 meters ahead of the vehicle in the direction in which the vehicle is traveling, the vehicle model unit 18 activates the brake. It is also possible to output that the vehicle will stop in front of meters. The simulation control unit 10 may display the simulation result by the vehicle model unit 18 on a display device, or may store it in a storage device.

FIG. 5 is a block diagram showing details of the sensor model unit 50 of the first embodiment. The sensor model unit 50 of the first embodiment includes a tag quantization unit 52, a true value quantization unit 54, an error derivation unit 56, a simulated value derivation unit 58, and a model storage unit 40. The model storage unit 40 corresponds to the storage unit described in the outline above. The model storage unit 40 stores sensor models generated by the sensor model generation device 20 .

The tag quantization unit 52 receives the tag data generated by the environment data generation unit 14 via the simulation control unit 10 and classifies it into one of a plurality of categories of predetermined explanatory variables. For example, the tag quantization unit 52 may classify the tag data indicating the weather into the corresponding category among the three categories, as shown in FIG.

The true value quantization unit 54 receives the true value data generated by the environment data generation unit 14 via the simulation control unit 10 and classifies it into one of a plurality of categories of predetermined explanatory variables. For example, the true value quantization unit 54 may classify the true value of the distance from the vehicle to the pedestrian into the corresponding category among ten categories, as shown in FIG.

The error derivation unit 56 derives the detection distance error according to the true value data and the tag data generated by the environment data generation unit 14 and the sensor model (for example, Equation 1 above) stored in the model storage unit 40. . In the embodiment, the error derivation unit 56 sets the value of the dummy variable corresponding to the category selected by the tag quantization unit 52 and the true value quantization unit 54 among the dummy variables of the sensor model to 1 (for non-selected categories By setting the corresponding dummy variable value to 0), the detected distance error, which is the objective variable of the sensor model, is calculated.

The simulated value derivation unit 58 adds the detection distance error derived by the error derivation unit 56 to the true value data generated by the environment data generation unit 14, thereby obtaining the detection result by the image analysis unit 16 (image analysis unit 28). to simulate. For example, the simulated value derivation unit 58 calculates the distance to the pedestrian that the image analysis unit 16 would detect if there was a pedestrian 6 meters ahead of the actual vehicle (a value including an error, such as 6.2 meters). A value derived by the simulated value derivation unit 58 (for example, the distance to a pedestrian) is input to the vehicle model unit 18 via the simulation control unit 10, and the behavior of the vehicle is simulated.

The operation of the above configuration will be described.
First, the operation of the sensor model generation device 20 of the first embodiment will be described with reference to FIG. Developers of in-vehicle equipment who should provide a vehicle simulation system to customers and partner companies actually run the vehicle, capture pedestrians outside the vehicle with an in-vehicle camera, and store multiple sample captured images in the sensor data storage unit 22. Memorize. At the same time, the developer causes the LIDAR device to measure the distance to the pedestrian, and causes the reference data storage unit 24 to store the measurement results indicating the true value of the distance of a plurality of samples. Furthermore, the developer causes the tag data storage unit 26 to store a plurality of samples of tag data indicating qualitative content such as weather.

The image analysis unit 28 detects the distance to the pedestrian from the image stored in the sensor data storage unit 22 for each sample. The true value derivation unit 30 detects the true value of the distance to the pedestrian from the measurement results stored in the reference data storage unit 24 for each sample. The error derivation unit 32 derives a detection distance error, which is the difference between the detection distance based on the image and the true value, for each sample. The true value quantization unit 34 identifies the true value category of the distance to the pedestrian, and the tag quantization unit 36 identifies the tag data category.

The model generation unit 38 uses the category of the true value of the distance to the pedestrian and the category of the tag data as dummy variables, and the detection distance error as the objective variable. Generate a model. The model generation unit 38 stores the generated sensor model in the model storage unit 40 .

According to the sensor model generation device 20 of the first embodiment, when the detection result by image analysis is required when simulating the behavior of the vehicle, the image of each case to be simulated is prepared, and the image of each case is prepared. Analysis becomes unnecessary, and the time and cost required for simulation can be reduced. In addition, by generating a sensor model using data representing the actual state of the surroundings of the vehicle, it is possible to generate a sensor model that outputs realistic results.

Next, the operation of the vehicle simulation system 110 of the first embodiment will be described with reference to FIGS. 4 and 5. FIG. Here, the behavior of the vehicle is simulated when there is a pedestrian in the vicinity of the vehicle. The user inputs the distance from the vehicle to the pedestrian and the weather as simulation parameters into the user interface 12 of the vehicle simulation system 110 . The environment data generation unit 14 generates true value data received by the user interface 12 indicating the distance from the vehicle to the pedestrian, and generates tag data received by the user interface 12 indicating the weather.

The tag quantization unit 52 identifies the category of tag data (eg weather in FIG. 2), and the true value quantization unit 54 identifies the category of true value data (eg distance in FIG. 2). The error deriving unit 56 extracts the categories (dummy variables) specified by the tag quantization unit 52 and the true value quantization unit 54 among the dummy variables indicating the categories of the explanatory variables of the sensor model stored in the model storage unit 40. By setting the value to "1", the detection distance error, which is the objective variable, is obtained. The simulated value derivation unit 58 adds the detection distance error to the distance (true value) from the vehicle to the pedestrian, thereby estimating the distance (including error) from the vehicle to the pedestrian detected by the image analysis unit 28. .

The vehicle model unit 18 determines the behavior of the vehicle based on the distance (including error) from the vehicle to the pedestrian derived by the simulated value deriving unit 58 . The simulation control unit 10 outputs the content determined by the vehicle model unit 18 as a simulation result to a predetermined output device or stores it in a predetermined storage device.

According to the vehicle simulation system 110 of the first embodiment, when a detection result by image analysis is required when simulating the behavior of a vehicle, it is possible to prepare an image of each case to be simulated and analyze the image of each case. It becomes unnecessary, and the time and cost required for simulation can be reduced. In addition, it becomes easier to increase the comprehensiveness of the simulation. Furthermore, by using a sensor model generated using the actual conditions around the vehicle, it is possible to simulate realistic vehicle behavior and the like.

The present disclosure has been described above based on the first embodiment. It should be understood by those skilled in the art that the first embodiment is an example, and that various modifications can be made to the combination of each component or each treatment process, and such modifications are within the scope of the present disclosure. Modifications are shown below.

A first modified example will be described. The sensor model unit 50 of the vehicle simulation system 110 applies a coefficient corresponding to the true value of the distance to an object (for example, a pedestrian) to the sensor model by curve fitting from a plurality of coefficients related to a plurality of dummy variables. .

FIG. 6 shows an example of curve fitting. Here, it is assumed that the sensor model includes at least 58 categories (explanatory variables). For example, the distance between the vehicle and the pedestrian may be set as a category in increments of 10 centimeters. The category coefficient graph 60 is a line graph showing coefficients (category coefficients) of 58 categories (explanatory variables) in the sensor model. The error derivation unit 56 of the sensor model unit 50 derives an approximated curve 62 (also referred to as a second-order polynomial approximated curve) by performing curve fitting on a plurality of category coefficients.

The error derivation unit 56 acquires from the approximated curve 62 category coefficients corresponding to the true value of the distance from the vehicle to the pedestrian. For example, when the distance is in increments of 10 centimeters, if the true value of the distance is 3.2 meters, the error derivation unit 56 uses the value (for example, "0") of the approximated curve 62 in the category number "32" as the category coefficient. may be obtained. The error derivation unit 56 applies the category coefficient obtained from the approximate curve 62 as a weight of the dummy variable (the value of the dummy variable is “1” in the example of FIG. 2), and derives the value of the objective variable (for example, the detection distance error). You may Alternatively, the true distance value may be converted to the coordinates of the horizontal axis as a continuous quantity, and the value calculated by substituting it into the approximate curve formula may be applied as the coefficient corresponding to the true distance value. According to these modifications, curve fitting can achieve appropriate weighting of the input explanatory variables.

A second modification will be described. The user interface 12 of the vehicle simulation system 110 may function as a reception unit that receives designation of data related to random numbers from the user. The simulated value derivation unit 58 of the sensor model unit 50 may add white noise to the estimated value using the sensor model based on the data on the random number specified by the user.

For example, the user may input a seed for generating a pseudo-random number sequence into the user interface 12 as data related to random numbers. The simulated value deriving unit 58 generates a pseudorandom number sequence based on the seed input by the user, generates white noise data based on the pseudorandom number sequence, and estimates the distance between the vehicle and the pedestrian. White noise may be added to the values. Since the data of the real environment contains a white noise component, according to this modified example, it is possible to add the white noise and obtain a more realistic estimation result.

A third modification will be described. Although the distance between the vehicle and the pedestrian is estimated by the sensor model in the above embodiment, the technique described in the embodiment can be applied to the case of estimating various conditions around the vehicle by the model. The third modified example shows an example of estimating the position of a parking frame (which can also be called a parking space) using a sensor model.

FIG. 7 shows the state around the vehicle in the modified example. In this modification, the XY coordinates of the four corners (FL, FR, BL, BR) of the parking frame 76 and the direction of the parking frame 76 (the angle of the parking frame 76 with respect to the traveling direction of the vehicle, θ in FIG. 7) Represented by In this modification, the coordinates of each of the four corners of the parking frame 76 detected by the image analysis unit are the distance from the camera 72 of the four corners (FL, FR, BL, BR) (for example, the distance of the front right point FR range _R in the case), the angle of the camera 72 with the optical axis 74 (eg, β _R in the case of the front right point FR), and the direction of the parking frame 76 (θ). The sensor model generating device 20 of this modified example categorizes the distance from the camera 72 , the angle with the optical axis 74 of the camera 72 , and the direction of the parking frame 76 .

Specifically, the image analysis unit 28 of the sensor model generation device 20 detects the position (coordinates, etc.) of the parking frame as the state of the surroundings of the vehicle from the image data showing the surroundings of the vehicle. The true value derivation unit 30 detects the true value of the position (coordinates, etc.) of the parking frame based on the detection result of the LIDAR device. The model generation unit 38 calculates the coordinates of the parking frame 76 detected by the true value deriving unit 30 (the coordinates of FR), the coordinates of the parking frame 76 detected by the image analysis unit 28 (the coordinates of FR), and is identified as the objective variable.

The model generator 38 also treats the distance from the camera 72, the angle of the camera 72 with respect to the optical axis 74, and the angle of the direction θ of the parking frame 76 with respect to the camera's optical axis 74 as explanatory variables. These data are collected by tests using actual vehicles, and a plurality of samples are created, as in the example. The model generator 38 performs multiple regression analysis based on multiple samples to obtain coefficients for each category (dummy variable). Incidentally, in practice, a sensor model corresponding to each of the coordinates of the four corners of the parking frame 76 may be generated.

The simulation control unit 10 of the vehicle simulation system 110 receives the coordinates of the four corners (FL, FR, BL, BR) and the direction of the parking frame from the user or the like as simulation parameters (the coordinate system may be the vehicle coordinate system, the sensor coordinates system), and inputs those parameters to the sensor model unit 50 . The sensor model unit 50 estimates the detection result by the image analysis unit 28, that is, estimates the coordinates of the parking frame 76 by inputting simulation parameters into the sensor model. Specifically, the sensor model unit 50 first converts the input coordinates of the four corners (FL, FR, BL, BR) and the direction of the parking frame into the distance from the camera 72 and the light of the camera 72, which are explanatory variables. Transform the angle with the axis 74 and the direction of the parking frame 76 . Next, the sensor model unit 50 obtains categories based on the values of the explanatory variables obtained, and applies coefficients corresponding to them. The vehicle model unit 18 simulates automatic driving of the vehicle (for example, automatic parking into the parking frame 76 ) based on the coordinates of the parking frame 76 estimated by the sensor model unit 50 .

<Second embodiment>
Vehicle-related simulations may be performed based on the amount of movement of the vehicle (hereinafter also referred to as "odometry"). The amount of computation based on the theoretical model of odometry is not large. However, due to environmental factors, odometry modeling is not easy, and the difference (that is, error) between the model output value and the true value tends to increase. The environmental factors include, for example, the degree of slippage between the tires and the road surface, which depends on the weather, road surface material, vehicle weight, and the like.

Therefore, the sensor model generation device of the second embodiment generates a sensor model by statistical processing using quantification analysis based on environmental data indicating the actual state of the vehicle or its surroundings and data on the amount of movement of the vehicle. Generate. The vehicle simulation system of the second embodiment realizes highly accurate simulation by utilizing the sensor model. Among the constituent elements of the sensor model generation device and vehicle simulation system of the second embodiment, the same reference numerals are given to those that are the same as or correspond to the members described in the first embodiment. In addition, repetitive descriptions of the contents that overlap with the first embodiment will be omitted as appropriate.

First, the sensor model of the second embodiment will be explained. The sensor model of the second embodiment is a model for estimating the movement amount of a vehicle in a vehicle simulation system. Specifically, the sensor model of the second embodiment uses the component of the movement amount detected by the movement amount derivation unit 120 described later as an explanatory variable, and the movement amount component detected by the true value derivation unit 30 described later. This model uses the difference between the true value of the component and the component value of the movement amount detected by the movement amount derivation unit 120 as the objective variable. The components of the amount of movement include the amount of lateral movement, the amount of vertical movement, and the turning angle.

FIG. 8 shows an example of sample data for constructing a sensor model in the second embodiment, that is, sample data for constructing a sensor model relating to the amount of movement of the vehicle. FIG. 8 shows some of the factors of the sample data (that is, explanatory variables of the sensor model). The explanatory variables of the sensor model in the second embodiment are (1) lateral movement amount, (2) vertical movement amount, (3) turning angle, (4) weather, (5) temperature, (6) vehicle weight, (7) ) tire type, and (8) road surface type.

The amount of lateral movement is the amount by which the vehicle moves in the lateral direction (that is, in the direction perpendicular to the direction of travel) per predetermined unit time, in other words, it is the speed of movement in the lateral direction. The vertical movement amount is the amount of movement of the vehicle in the vertical direction (that is, the direction of travel) per predetermined unit time, in other words, the speed of movement in the vertical direction. The amount of lateral movement and the amount of longitudinal movement may be categorized in units of, for example, 10 kilometers per hour, but preferably in smaller units at lower speeds. For example, speeds less than 15 kilometers per hour may be categorized by 5 kilometers per hour. The turning angle is the angle by which the vehicle turns per predetermined unit time, in other words, the angular velocity. Turn angles may be categorized in 0.5 degree increments, for example.

The weather is the weather when the vehicle is running, and includes sunny, cloudy, and rainy, as in the first embodiment. The temperature is the temperature when the vehicle is running, and may be categorized in units of 2° C., for example. Vehicle weight may, for example, be categorized in tens of kilograms (eg, the average weight of an adult or child). The tire type is the type of tire mounted on the vehicle, and may be categorized by summer tire or winter tire, for example. The road surface type is the type of road surface on which the vehicle travels, and may be categorized by, for example, asphalt, dirt, slope, and the like. As already mentioned, each category of factor is also called a dummy variable.

In addition, as data indicating the accuracy of movement amount detection, each sample is the difference between the true value of the movement amount of the vehicle and the movement amount of the vehicle detected based on the sensor data (hereinafter also referred to as "movement amount error"). .)including. In the second embodiment, the true value of the movement amount of the vehicle is obtained by positioning processing using a GNSS (Global Navigation Satellite System) such as a GPS (Global Positioning System).

By performing multiple regression analysis with each category of each factor as a dummy variable (explanatory variable) and the movement distance error as the objective variable, the magnitude of the influence of each category of each factor on the movement distance error (hereinafter referred to as "category coefficient") ) can be generated. If the number of factors is m, the number of categories of each factor is _{n 1} , . It can be represented by 1. As already mentioned, a ^ij is the category coefficient on the dummy variable (X _ijk ) representing the jth category of the ith factor.

FIG. 9 is a block diagram showing functional blocks of the sensor model generation device 20 of the second embodiment. The sensor model generation device 20 of the second embodiment includes a sensor data storage unit 22, a reference data storage unit 24, a tag data storage unit 26, a movement amount derivation unit 120, a true value derivation unit 30, an error derivation unit 32, a movement amount quantum It has a quantization unit 121 , a tag quantization unit 36 and a model generation unit 38 .

The sensor data storage unit 22 stores sample data indicating output results from various onboard sensors such as a speed sensor and a steering angle sensor, in other words, sample data indicating results of detection of the actual vehicle state. In the embodiment, the sensor data storage unit 22 stores at least the vehicle speed and steering angle (or turning angle) for each sample.

The reference data storage unit 24 stores the true value of the actual vehicle state. In the embodiment, the reference data storage unit 24 stores, for each sample, data in which the current position of the vehicle specified by the GPS device is arranged in time series, in other words, data indicating the transition of the vehicle position in time series. .

The tag data storage unit 26 stores tag data, which is data relating to the actual state of the vehicle obtained by means other than sensors. The tag data storage unit 26 stores tag data set by users (developers, testers, etc.). The tag data may include, for example, the weather, temperature, vehicle weight, tire type, and road surface type when a vehicle test or the like (that is, collection of sample data) is performed.

The data stored in the sensor model generation device 20, the sensor data storage unit 22, and the reference data storage unit 24, which are collected or input in the same test, are associated with each other to form a set of sample data. . A plurality of sets of sample data (for example, 10,000 sets of sample data as shown in FIG. 8) are stored in the sensor model generation device 20, the sensor data storage unit 22, and the reference data storage unit 24. FIG.

The movement amount derivation unit 120 corresponds to the first derivation unit described in the outline above. Based on the sample data indicating the output results of the sensors stored in the sensor data storage unit 22, the movement amount deriving unit 120 calculates the components of the movement amount of the vehicle as the state of attributes related to automatic driving of the vehicle for each sample. detect. Specifically, the movement amount derivation unit 120 calculates the lateral movement amount, the vertical movement amount, and the turning angle (all of which are values per predetermined unit time) of each sample based on the speed and steering angle indicated by each sample. derive A known technique may be adopted for this derivation method.

The true value derivation unit 30 corresponds to the second derivation unit described in the outline above. The true value derivation unit 30 detects the component of the movement amount of the vehicle for each sample based on the time-series position transition of the vehicle indicated by each sample stored in the reference data storage unit 24 . Specifically, the true value derivation unit 30 derives the lateral movement amount, the vertical movement amount, and the turning angle (all values per predetermined unit time) of each sample. A well-known technique may also be adopted for this derivation method.

Note that the true value derivation unit 30 derives the components of the amount of movement of the vehicle based on changes in the actual vehicle position. Therefore, the component value of the vehicle movement amount derived by the true value derivation unit 30 has higher precision than the component value of the vehicle movement amount derived by the movement amount derivation unit 120 . In the second embodiment, the component value of the amount of movement of the vehicle derived by the true value derivation unit 30 is treated as the true value. Strictly speaking, the component value of the amount of movement of the vehicle derived by the true value derivation unit 30 may be different from the true value, but it is within a range (within a predetermined threshold value) that can be considered to be the same as the true value. is desirable.

The error derivation unit 32 derives the difference between the detection result by the true value derivation unit 30 and the detection result by the movement amount derivation unit 120 for each sample and for each type of movement amount component. That is, the error derivation unit 32 derives a lateral movement amount error, a vertical movement amount error, and a turning angle error (hereinafter collectively referred to as "movement amount error") for each sample. For example, FIG. 8 shows an example in which the objective variable is the lateral movement amount error. Three patterns are generated, one for vertical displacement error and the other for turning angle error as the objective variable.

The movement amount quantization unit 121 classifies the detection result by the movement amount derivation unit 120 into one of a plurality of categories of predetermined explanatory variables for each sample. For example, the movement amount quantization unit 121 classifies each of the lateral movement amount, the vertical movement amount, and the turning angle detected by the movement amount derivation unit 120 into one of a plurality of predetermined categories. In the second embodiment, as shown in FIG. 8, the value of the applicable category is set to "1".

The tag quantization unit 36 classifies the tag data stored in the tag data storage unit 26 into one of a plurality of categories of predetermined explanatory variables for each sample. For example, as shown in FIG. 8, the tag quantization unit 36 classifies the tag data indicating the weather into the corresponding category among the three categories, and sets the value of the corresponding category to "1".

The model generator 38 corresponds to the generator described in the outline above. The model generation unit 38 is a sensor model for estimating the detection result by the true value derivation unit 30, that is, it generates a sensor model for estimating the true value of the amount of movement of the vehicle. The model generation unit 38 uses the components of the vehicle movement amount derived by the movement amount derivation unit 120 as explanatory variables, and the component values (that is, the true values) of the vehicle movement amounts derived by the true value derivation unit 30. , a sensor model whose objective variable is the difference between the component values of the amount of movement of the vehicle derived by the movement amount derivation unit 120 .

Specifically, as illustrated in FIG. 8, the model generation unit 38 generates the error derived by the error derivation unit 32, the category selected by the movement amount quantization unit 121, and the category selected by the tag quantization unit 36. A combination with the selected category is created for each sample and for each movement amount component. The model generation unit 38 uses the error calculated by the error derivation unit 32 as the objective variable, and the category selected by the movement amount quantization unit 121 and the category selected by the tag quantization unit 36 as dummy variables (values are "1"), a plurality of category coefficients for a plurality of categories are derived by executing statistical processing (multiple regression analysis in the embodiment).

The model generating unit 38 generates, as a sensor model, a regression equation (equation 1 above) in which category coefficients are set for each component of the movement amount, and stores data of the sensor model in the model storage unit 40 . That is, the model generator 38 generates a sensor model whose objective variable is the error in the lateral displacement, a sensor model whose objective variable is the error in the longitudinal displacement, and a sensor model whose objective variable is the error in the turning angle. Stored in the model storage unit 40 .

Next, a vehicle simulation system using the sensor model generated by the sensor model generation device 20 will be described. The basic functional blocks of the vehicle simulation system 110 of the second embodiment are the same as those of the vehicle simulation system 110 of the first embodiment shown in FIG.

The environmental data generator 14 generates environmental data including explanatory variables of the sensor model. The environmental data of the second embodiment includes, as vehicle simulation parameters, assumed values of the amount of movement of the vehicle (horizontal movement amount, vertical movement amount, and turning angle per unit time) based on the output results from the vehicle-mounted sensors. This assumed value can also be said to be an assumed value of the movement amount component detected by the movement amount derivation unit 120 . The environmental data of the second embodiment further includes tag data as vehicle simulation parameters, such as weather, temperature, vehicle weight, and the like.

The sensor model unit 50 corresponds to the estimation unit described in the outline above. The sensor model unit 50
By inputting the movement amount component of the vehicle detected by the movement amount derivation unit 120 into the sensor model, the true value of the movement amount component of the vehicle is estimated.

The vehicle model section 18 corresponds to the simulation section described in the outline above. The vehicle model unit 18 executes vehicle simulation based on the true value of the vehicle movement amount component estimated by the sensor model unit 50 . For example, the vehicle model unit 18 may determine the next behavior of the vehicle based on the true value of the movement amount component of the vehicle.

The simulation control unit 10 may display the simulation result by the vehicle model unit 18 on a display device (not shown) or store it in a storage device (not shown). The person in charge of the simulation may confirm the content of the next behavior determined by the vehicle model section 18 on the display device, and confirm whether or not there is any problem with the next behavior.

FIG. 10 is a block diagram showing details of the sensor model unit 50 of the second embodiment. The sensor model unit 50 of the second embodiment includes a tag quantization unit 52, a movement amount quantization unit 122, an error derivation unit 56, a correction unit 124, and a model storage unit 40. The model storage unit 40 corresponds to the storage unit described in the outline above. The model storage unit 40 stores the sensor model generated by the sensor model generation device 20 for each movement amount component of the vehicle.

The movement amount quantization unit 122 receives the movement amount component values generated by the environment data generation unit 14 via the simulation control unit 10, and classifies each component into one of a plurality of categories of predetermined explanatory variables. Classify. For example, as shown in FIG. 8, the movement amount quantization unit 122 classifies the lateral movement amount among the movement amount component values into the corresponding category among a plurality of categories (0 to 5, 5 to 10, etc.). .

The error derivation unit 56 derives the movement amount error according to the movement amount component value and the tag data generated by the environment data generation unit 14 and the sensor model (for example, Equation 1 above) stored in the model storage unit 40. do. Specifically, the error derivation unit 56 calculates the dummy variables corresponding to the category selected (that is, determined to be applicable) by the tag quantization unit 52 and the movement amount quantization unit 122 among the dummy variables of the sensor model. By setting the value to 1, the value of the displacement error, which is the objective variable of the sensor model, is calculated. Note that the error deriving unit 56 uses a sensor model for the error in the lateral movement amount, a sensor model for the error in the vertical movement amount, and a sensor model for the error in the turning angle to calculate the error in the lateral movement amount, the error in the vertical movement amount, Derive each of the turning angle errors.

The correction unit 124 adds the error for each vehicle movement amount component derived by the error deriving unit 56 to the movement amount component value generated by the environment data generation unit 14, thereby obtaining the true value for each vehicle movement amount component. to derive The true value for each vehicle movement amount component derived by the correction unit 124 corresponds to the detection result of the vehicle movement amount by the true value derivation unit 30 having higher accuracy than the movement amount derivation unit 120 . The values derived by the correction unit 124 (that is, the lateral movement amount, the vertical movement amount, and the turning angle as true values) are input to the vehicle model unit 18 via the simulation control unit 10, and the behavior of the vehicle, for example, is simulated. .

The operation of the above configuration will be described.
First, the operation of the sensor model generation device 20 of the second embodiment will be described with reference to FIG. Developers of in-vehicle equipment who should provide a vehicle simulation system to their customers and partner companies should actually run the vehicle and collect multiple samples of sensor data output from the in-vehicle sensors (speed sensor, steering angle sensor, etc.). It is stored in the data storage unit 22 . Each in-vehicle sensor may output sensor data indicating detection results to a CAN (Controller Area Network), which is an in-vehicle network, and the sensor data storage unit 22 may collect and store sensor data flowing through the CAN. .

At the same time, the developer causes the reference data storage unit 24 to store a plurality of samples of position data (data indicating changes in the position of the vehicle) measured by the GPS device while the vehicle is running. Furthermore, the developer causes the tag data storage unit 26 to store a plurality of samples of tag data including qualitative information such as weather and road surface type.

The movement amount derivation unit 120 derives the value of the movement amount component (horizontal movement amount, vertical movement amount, turning angle) of the vehicle based on the sensor data stored in the sensor data storage unit 22 for each sample. The true value derivation unit 30 derives the true value of the movement amount component of the vehicle from the position data stored in the reference data storage unit 24 for each sample. Specifically, the true value derivation unit 30 reads out the position data of the current frame and the position data of the previous frame stored in the reference data storage unit 24, and calculates the true value of the movement amount component of the vehicle according to the difference. Calculate the value. The error deriving unit 32 derives a lateral movement amount error, a vertical movement amount error, and a turning angle error for each sample. The movement amount quantization unit 121 identifies the category of the value of the movement amount component detected by the movement amount derivation unit 120 for each sample. The tag quantization unit 36 identifies the category of tag data for each sample.

The model generation unit 38 uses the category of the value of the movement amount component detected by the movement amount derivation unit 120 and the category of the tag data as dummy variables, and the error of the movement amount component of the vehicle as the objective variable. A sensor model is generated by performing multiple regression analysis. In the second embodiment, the model generation unit 38 generates three types of sensor models, each of which has a lateral movement amount error, a vertical movement amount error, and a turning angle error as objective variables. The model generator 38 stores three types of sensor models in the model storage 40 .

According to the sensor model generation device 20 of the second embodiment, it is possible to realize a sensor model relating to the amount of movement of the vehicle in consideration of environmental factors, and outputting a realistic result. For example, it is possible to realize a sensor model that outputs results reflecting qualitative information such as weather and road surface type. Further, in the sensor model of the second embodiment, it is possible to obtain data on the true value of the movement amount from the movement amount value that may contain an error based on sensor detection.

Next, the operation of the vehicle simulation system 110 of the second embodiment will be described with reference to FIGS. 4 and 10. FIG. The environment data generation unit 14 generates movement amount component values and tag data as parameters for vehicle simulation in accordance with user operations input via the user interface 12 . The values of these parameters may be specified by the user (user terminal). Further, the movement amount component values as parameters of the vehicle simulation are assumed values of the lateral movement amount, the vertical movement amount, and the turning angle based on the output results from the vehicle-mounted sensors, and may differ from the true values.

The tag quantization unit 52 identifies categories of tag data (for example, weather, road surface type, etc.) generated by the environment data generation unit 14 . The movement amount quantization unit 122 identifies categories of movement amount component values generated by the environment data generation unit 14 .

The error deriving unit 56 extracts the category (dummy variable) to "1". As a result, the error derivation unit 56 obtains the lateral movement amount error, the vertical movement amount error, and the turning angle error as the objective variable values of the three types of sensor models. The correction unit 124 adds an error to the movement amount component value generated by the environment data generation unit 14 to estimate the true value of the lateral movement amount, the true value of the vertical movement amount, and the true value of the turning angle.

The vehicle model unit 18 executes a vehicle simulation based on the true value of the lateral movement amount, the true value of the vertical movement amount, and the true value of the turning angle estimated by the correction unit 124, and, for example, determines the following behavior of the vehicle. decide. The simulation control unit 10 outputs the result of the simulation by the vehicle model unit 18 to a predetermined output device or stores it in a predetermined storage device.

According to the vehicle simulation system 110 of the second embodiment, a highly accurate simulation can be realized by using a sensor model relating to the amount of movement of the vehicle that takes environmental factors into account, and that uses a sensor model that outputs realistic results. . For example, it is possible to accurately estimate the true value of the amount of movement that takes environmental factors into consideration based on the assumed value of the amount of movement based on the output result from the vehicle-mounted sensor.

The present disclosure has been described above based on the second embodiment. The second embodiment is an example, and those skilled in the art will understand that various modifications can be made to the combination of each component or each treatment process, and such modifications are within the scope of the present disclosure.

The curve fitting described in the first modification of the first embodiment can also be applied as a modification of the second embodiment. An estimating unit (corresponding to the error deriving unit 56 in the second embodiment) that obtains the value of the objective variable using the sensor model obtains continuous explanatory variables by curve fitting from each category coefficient related to a plurality of dummy variables. A continuous value of the categorical coefficients for the values may be determined. The error derivation unit 56 may apply, to the sensor model, a category coefficient value corresponding to the explanatory variable value in the vehicle simulation from among the continuous values of the category coefficient.

As in the example of FIG. 6, the error deriving unit 56 may derive an approximate curve by performing curve fitting on a plurality of category coefficients corresponding to a plurality of categories of lateral displacement. The error derivation unit 56 may similarly derive approximate curves corresponding to a plurality of category coefficients for quantitative data (air temperature, etc.) among the vertical movement amount, turning angle, and tag data. The error derivation unit 56 acquires category coefficients corresponding to the quantitative values of the lateral movement amount, the vertical movement amount, the turning angle, and the tag data generated by the environment data generation unit 14 from the approximate curve. , may be fitted to the sensor model (eg, Equation 1 above). Setting the value of the dummy variable to "1" is the same as in the embodiment. According to this modification, it is possible to appropriately weight the input explanatory variable values by curve fitting.

The addition of white noise described in the second modification of the first embodiment can also be applied as a modification of the second embodiment. In other words, the user interface 12 of the vehicle simulation system 110 may function as a reception unit that receives designation of data regarding random numbers from the user. The correction unit 124 of the sensor model unit 50 may add white noise to the estimated value (error value or movement amount component value) using the sensor model based on data related to random numbers specified by the user. According to this modified example, it is possible to obtain a more realistic estimation result by adding white noise.

<Third embodiment>
In vehicle simulation, there are cases where the exact value (true value) of the amount of movement of the vehicle is known based on the scenario. In this case, the sensor data (vehicle speed, steering angle, etc.) can be calculated by known calculations based on the true value of the amount of movement of the vehicle. becomes something different.

Similarly to the second embodiment, the sensor model generation device of the third embodiment also uses quantification analysis based on environmental data indicating the actual state of the vehicle or its surroundings and data on the amount of movement of the vehicle. A sensor model is generated by statistical processing. The vehicle simulation system of the third embodiment realizes highly accurate simulation by utilizing the sensor model. Among the members constituting the sensor model generation device and the vehicle simulation system of the third embodiment, those that are the same as or correspond to the members described in the first or second embodiment are denoted by the same reference numerals. In addition, re-explanation of the same content as in the first embodiment or the second embodiment will be omitted as appropriate.

First, the sensor model of the third embodiment will be explained. The sensor model of the third embodiment is a model for estimating the movement amount of a vehicle in a vehicle simulation system. Specifically, in the sensor model of the third embodiment, the moving amount component detected by the true value deriving section 30 is used as an explanatory variable, and the true value of the moving amount component detected by the true value deriving section 30 is , and the component value of the movement amount detected by the movement amount derivation unit 120 as the objective variable. The sensor model of the third embodiment can also be expressed by Equation 1 above, as in the first and second embodiments.

Sample data for constructing the sensor model in the third embodiment is the same as in FIG. However, in the second embodiment, at least (1) the amount of lateral movement, (2) the amount of vertical movement, and (3) the turning angle are derived by the movement amount derivation unit 120, and are values that may differ from the true values. there were. On the other hand, in the third embodiment, at least (1) the amount of lateral movement, (2) the amount of vertical movement, and (3) the turning angle are derived by the true value deriving section 30, and value).

FIG. 11 is a block diagram showing functional blocks of the sensor model generation device 20 of the third embodiment. The sensor model generation device 20 of the third embodiment includes a true value quantization section 34 instead of the displacement quantization section 121 in the sensor model generation device 20 of the second embodiment. Of the functional blocks in the sensor model generation device 20 of the third embodiment, a sensor data storage unit 22, a reference data storage unit 24, a tag data storage unit 26, a movement amount derivation unit 120, a true value derivation unit 30, and an error derivation unit 32 , and the tag quantization unit 36 are the same as those in the second embodiment, so description thereof will be omitted.

The true value quantization unit 34 classifies the detection result by the true value derivation unit 30 into one of a plurality of categories of predetermined explanatory variables for each sample. For example, the true value quantization unit 34 calculates a predetermined plurality of values for each of the lateral movement amount, the vertical movement amount, and the turning angle (all of which are true values per predetermined unit time) detected by the true value deriving unit 30. classified into one of the following categories. In the third embodiment, as shown in FIG. 8, the value of the applicable category is set to "1".

The model generation unit 38 is a sensor model for estimating the detection result by the movement amount derivation unit 120, in other words, it generates a sensor model for estimating sensor data output from the vehicle-mounted sensor. The model generation unit 38 uses the components of the vehicle movement amount derived by the true value derivation unit 30 as explanatory variables, and the component values (i.e., true values) of the vehicle movement amounts derived by the true value derivation unit 30 as , a sensor model whose objective variable is the difference between the component values of the amount of movement of the vehicle derived by the movement amount derivation unit 120 .

Specifically, as illustrated in FIG. 8, the model generation unit 38 generates the error derived by the error derivation unit 32, the category selected by the true value quantization unit 34, and the category selected by the tag quantization unit 36. A combination with the selected category is created for each sample and for each movement amount component. The model generation unit 38 uses the error calculated by the error derivation unit 32 as the objective variable, and the category selected by the true value quantization unit 34 and the category selected by the tag quantization unit 36 as dummy variables (values are "1"), a plurality of category coefficients for a plurality of categories are derived by executing statistical processing (multiple regression analysis in the embodiment).

The model generating unit 38 generates, as a sensor model, a regression equation (equation 1 above) in which category coefficients are set for each component of the movement amount, and stores data of the sensor model in the model storage unit 40 . That is, the model generator 38 generates a sensor model whose objective variable is the error in the lateral displacement, a sensor model whose objective variable is the error in the longitudinal displacement, and a sensor model whose objective variable is the error in the turning angle. Stored in the model storage unit 40 .

Next, a vehicle simulation system using the sensor model generated by the sensor model generation device 20 will be described. Basic functional blocks of the vehicle simulation system 110 of the third embodiment are similar to those of the vehicle simulation system 110 of the first embodiment shown in FIG.

The environmental data generator 14 generates environmental data including explanatory variables of the sensor model. The environment data of the third embodiment is the true value of the movement amount component of the vehicle (horizontal movement amount per unit time, vertical movement amount, turning angle). This true value corresponds to the movement amount component value detected by the true value derivation unit 30 . Further, the environment data of the third embodiment further includes tag data as parameters for vehicle simulation, such as weather, temperature, vehicle weight, etc., as in the second embodiment.

The sensor model unit 50 is a value corresponding to the detection result by the true value derivation unit 30. That is, by inputting the true value of the movement amount component of the vehicle into the sensor model, the vehicle detected by the movement amount derivation unit 120 is calculated. Estimate the displacement component value of . In other words, the sensor model unit 50 calculates based on the output results of the vehicle-mounted sensors (speed sensor, steering angle sensor, etc.) when the vehicle moves as indicated by the true value (typically moves according to the simulation scenario). The movement amount of the vehicle detected by the sensor (hereinafter also referred to as "sensor-detected movement amount") is estimated. The sensor-detected movement amount includes the lateral movement amount, the vertical movement amount, and the turning angle per unit time, and may differ from the true value.

The vehicle model unit 18 executes vehicle simulation based on the sensor-detected movement amount estimated by the sensor model unit 50 . For example, the vehicle model unit 18 inputs the sensor-detected movement amount (estimated value) to an automatic driving control program (or an ECU (Electronic Control Unit) in which the program is implemented), and simulates the execution result of the automatic driving control program. You can output it as a result.

The simulation control unit 10 may display the simulation result by the vehicle model unit 18 on a display device (not shown) or store it in a storage device (not shown). The person in charge of the simulation may check the execution result of the automatic cruise control program output by the vehicle model unit 18 and check whether or not there is any problem.

FIG. 12 is a block diagram showing details of the sensor model unit 50 of the third embodiment. The sensor model unit 50 of the third embodiment includes a tag quantization unit 52, a true value quantization unit 54, an error derivation unit 56, a correction unit 124, and a model storage unit 40. The model storage unit 40 stores the sensor model generated by the sensor model generation device 20 for each movement amount component of the vehicle.

The true value quantization unit 54 receives the true value of the movement amount component generated by the environment data generation unit 14 via the simulation control unit 10, and selects one of a plurality of categories of predetermined explanatory variables for each component. classified as For example, as shown in FIG. 8, the true value quantization unit 54 categorizes the true value of the lateral movement amount among the true values of the movement amount components as corresponding to a plurality of categories (0 to 5, 5 to 10, etc.). Categorize into categories that

The error derivation unit 56 calculates the movement amount error according to the true value of the movement amount component and the tag data generated by the environment data generation unit 14 and the sensor model (for example, the above equation 1) stored in the model storage unit 40. to derive Specifically, the error derivation unit 56 sets the value of the dummy variable corresponding to the category selected by the tag quantization unit 52 and the true value quantization unit 54 among the dummy variables of the sensor model to 1. , the value of the displacement error, which is the objective variable of the sensor model, is calculated. Note that the error deriving unit 56 uses a sensor model for the error in the lateral movement amount, a sensor model for the error in the vertical movement amount, and a sensor model for the error in the turning angle to calculate the error in the lateral movement amount, the error in the vertical movement amount, Derive each of the turning angle errors.

The correction unit 124 adds the error for each movement amount component of the vehicle derived by the error derivation unit 56 to the true value of the movement amount component generated by the environment data generation unit 14, thereby obtaining each component of the sensor detected movement amount. to derive the value of The sensor-detected movement amount derived by the correction unit 124 corresponds to the detection result by the movement amount derivation unit 120 whose accuracy is lower than that of the true value derivation unit 30 . The value of each component of the sensor-detected movement amount derived by the correction unit 124 (that is, the amount of lateral movement, the amount of vertical movement, and the turning angle) is input to the vehicle model unit 18 via the simulation control unit 10, and is used for automatic driving, for example. The behavior of the controlling ECU is simulated.

The operation of the above configuration will be described.
First, the operation of the sensor model generation device 20 of the third embodiment will be described with reference to FIG. The operations of the movement amount derivation unit 120, the true value derivation unit 30, the error derivation unit 32, and the tag quantization unit 36 are the same as those in the second embodiment, and thus descriptions thereof will be omitted. The true value quantization unit 34 identifies the true value category of the movement amount component detected by the true value derivation unit 30 for each sample.

The model generation unit 38 uses the category of the true value of the movement amount component detected by the true value quantization unit 34 and the category of the tag data as dummy variables, and the error of the movement amount component as the objective variable. A sensor model is generated by performing multiple regression analysis. In the third embodiment, the model generation unit 38 also generates three types of sensor models, each of which has a lateral movement amount error, a vertical movement amount error, and a turning angle error as objective variables. The model generator 38 stores three types of sensor models in the model storage 40 .

According to the sensor model generating device 20 of the third embodiment, it is possible to realize a sensor model that is related to the amount of movement of the vehicle in consideration of environmental factors and that outputs results that are in line with reality. For example, it is possible to realize a sensor model that outputs results reflecting qualitative information such as weather and road surface type. Further, in the sensor model of the third embodiment, it is possible to obtain data on the sensor-detected movement amount that may contain an error from the true value of the movement amount of the vehicle.

Next, the operation of the vehicle simulation system 110 of the third embodiment will be described with reference to FIGS. 4 and 12. FIG. The environmental data generation unit 14 generates, as vehicle simulation parameters, the true value of the amount of movement determined by the user based on the simulation scenario and tag data in response to a user operation input via the user interface 12. . The tag quantization unit 52 of the sensor model unit 50 identifies categories of tag data (for example, weather, road surface type, etc.) generated by the environment data generation unit 14 . The true value quantization unit 54 identifies the true value category of the movement amount component generated by the environment data generation unit 14 .

The error deriving unit 56 extracts the category (dummy variable) to "1". As a result, the error derivation unit 56 obtains the lateral movement amount error, the vertical movement amount error, and the turning angle error as the objective variable values of the three types of sensor models. The correction unit 124 adds an error to the true value of the movement amount component generated by the environment data generation unit 14 to estimate the lateral movement amount, the vertical movement amount, and the turning angle as the movement amount detected by the sensor.

The vehicle model unit 18 executes vehicle simulation based on the lateral movement amount, the vertical movement amount, and the turning angle estimated by the correction unit 124 . For example, the vehicle model unit 18 inputs the lateral movement amount, the vertical movement amount, and the turning angle as sensor-detected movement amounts to a program (or ECU) for automatic driving control, and obtains the output result of the program (or ECU). . The vehicle model unit 18 derives corresponding sensor data (vehicle speed, steering angle, etc.) based on the amount of movement detected by the sensor, and inputs the sensor data to the automatic driving control program (or ECU). good too. The simulation control unit 10 outputs the result of the simulation by the vehicle model unit 18 to a predetermined output device or stores it in a predetermined storage device.

According to the vehicle simulation system 110 of the third embodiment, a highly accurate simulation can be realized by using a sensor model relating to the amount of movement of the vehicle that takes environmental factors into account, and that outputs realistic results. . For example, based on the true value of the amount of movement of the vehicle based on the simulation scenario, the amount of movement detected by the sensor when the actual vehicle behaves according to the scenario is specified, and the vehicle malfunctions at that amount of movement detected by the sensor. can confirm that it is not.

The present disclosure has been described above based on the third embodiment. The third embodiment is an example, and it should be understood by those skilled in the art that various modifications can be made to the combination of each component or each treatment process, and such modifications are within the scope of the present disclosure.

The curve fitting described in the first modification of the first embodiment and the addition of white noise described in the second modification of the first embodiment can be performed not only in the modification of the second embodiment but also in the third embodiment. Of course, it can also be applied as a modification of the example.

The techniques described in the second embodiment, the third embodiment, and the modification may be specified by the following items.
[Item 1]
a first derivation unit for deriving a state of an attribute related to automatic driving of the vehicle based on the output result of a sensor mounted on the vehicle;
a second derivation unit that derives the attribute state of the vehicle with higher accuracy than the first derivation unit;
A model for simulating the vehicle, wherein the state derived by the first derivation unit or the second derivation unit is used as an explanatory variable, and the state derived by the second derivation unit and the first derivation unit a generator that generates a model whose objective variable is the difference from the state derived by
A model generation device comprising:
[Item 2]
The state of the attributes related to automatic driving of the vehicle is a state of the vehicle surroundings for controlling the automatic driving of the vehicle,
The first derivation unit derives a state of the surroundings of the vehicle from image data showing the surroundings of the vehicle,
The model is a model for estimating the result of derivation by the first derivation unit in the simulation of the vehicle, wherein the state derived by the second derivation unit is used as an explanatory variable, and A model whose objective variable is the difference between the state derived by the first derivation unit and the state derived by the first derivation unit,
The model generation device according to item 1.
[Item 3]
The attribute state related to automatic driving of the vehicle is a component of the amount of movement of the vehicle,
The model is a model for estimating the result of derivation by the second derivation unit in the simulation of the vehicle, wherein the movement amount component derived by the first derivation unit is used as an explanatory variable, and the second derivation unit A model whose objective variable is the difference between the movement amount component derived by and the movement amount component derived by the first derivation unit,
The model generation device according to item 1.
[Item 4]
The attribute state related to automatic driving of the vehicle is a component of the amount of movement of the vehicle,
The model is a model for estimating the result derived by the first derivation unit in the simulation of the vehicle, and the component of the movement amount derived by the second derivation unit is used as an explanatory variable, and the second derivation unit A model whose objective variable is the difference between the movement amount component derived by and the movement amount component derived by the first derivation unit,
The model generation device according to item 1.
[Item 5]
With respect to a derivation unit that derives the state of an attribute related to automatic driving of the vehicle based on the output result of a sensor mounted on the vehicle, the state of the attribute derived by the derivation unit or the truth of the state of the attribute A model for estimating a value, wherein the state of the attribute derived by the derivation unit or the true value of the state of the attribute is used as an explanatory variable, and the true value of the state of the attribute and the true value of the attribute state by the derivation unit a storage unit that stores a model whose objective variable is the difference from the derived state of the attribute;
By inputting one of the state of the attribute derived by the derivation unit and the true value of the state of the attribute into the model as a parameter of the simulation, the state of the attribute derived by the derivation unit and the value of the attribute an estimating unit that estimates the other of the true values of the state;
a simulation unit that executes a simulation of the vehicle based on the estimation result of the estimation unit;
vehicle simulation system.
[Item 6]
The derivation unit derives the state of the vehicle periphery for controlling the automatic running of the vehicle as the attribute state related to the automatic running of the vehicle from the image data showing the periphery of the vehicle,
The model is a model for estimating the result of derivation by the derivation unit, wherein the true value of the state of the vehicle periphery is used as an explanatory variable, and the true value of the state of the vehicle periphery and the model derived by the derivation unit are It is a model with the difference from the state as the objective variable,
The estimation unit estimates the derivation result by the derivation unit by inputting the true value of the state around the vehicle as a simulation parameter into the model.
A vehicle simulation system according to item 5.
[Item 7]
The attribute state related to automatic driving of the vehicle is a component of the amount of movement of the vehicle,
The model is a model for estimating the true value of the movement amount component, wherein the movement amount component derived by the derivation unit is used as an explanatory variable, and the movement amount component true value and the derived It is a model whose objective variable is the difference between the component of the amount of movement derived by the part,
The estimation unit estimates the true value of the movement amount component by inputting the movement amount component derived by the derivation unit into the model.
A vehicle simulation system according to item 5.
[Item 8]
The attribute state related to automatic driving of the vehicle is a component of the amount of movement of the vehicle,
The model is a model for estimating the true value of the component of the movement amount, the true value of the component of the movement amount is used as an explanatory variable, and the true value of the component of the movement amount is derived by the derivation unit. is a model whose objective variable is the difference between the calculated movement amount component,
The estimation unit estimates the movement amount component derived by the derivation unit by inputting the true value of the movement amount component into the model.
A vehicle simulation system according to item 5.
[Item 9]
The model includes a plurality of dummy variables corresponding to multiple stages of values of the explanatory variable,
The estimating unit obtains a continuous value of the coefficient for the continuous value of the explanatory variable by curve fitting from each coefficient related to the plurality of dummy variables, and the value of the explanatory variable in the simulation from among the continuous values of the coefficient. applying to the model a coefficient value corresponding to
9. A vehicle simulation system according to any one of items 5-8.
[Item 10]
further comprising a reception unit that receives a specification of data related to the random number from the user,
The estimation unit adds white noise to the estimated value using the model based on the data on the random number.
A vehicle simulation system according to any one of items 5 to 9.
[Item 11]
A first derivation unit derives the attribute state related to automatic driving of the vehicle based on the output result of the sensor mounted on the vehicle,
A second derivation unit derives an attribute state related to the vehicle with higher accuracy than the first derivation unit,
A model for simulating the vehicle, wherein the state derived by the first derivation unit or the second derivation unit is used as an explanatory variable, and the state derived by the second derivation unit and the first derivation unit Generate a model whose objective variable is the difference from the state derived by
Model generation method.
[Item 12]
A first derivation unit derives the attribute state related to automatic driving of the vehicle based on the output result of the sensor mounted on the vehicle,
A second derivation unit derives an attribute state related to the vehicle with higher accuracy than the first derivation unit,
A model for simulating the vehicle, wherein the state derived by the first derivation unit or the second derivation unit is used as an explanatory variable, and the state derived by the second derivation unit and the first derivation unit Generate a model whose objective variable is the difference from the state derived by
A computer program that makes a computer do something.

<Fourth embodiment>
Regarding this embodiment, the following description will focus on the points that are different from the previous embodiments, and the description of the common points will be omitted. Among the constituent elements of this modified example, constituent elements that are the same as or correspond to the constituent elements of the previous embodiments are denoted by the same reference numerals.

In the conventional vehicle simulation system, only the vehicle position as the ideal position output by the vehicle model was drawn as the vehicle position of the simulation result. However, the vehicle position as the ideal position does not take into account side slipping of tires and the like, and may deviate from the actual vehicle position. Therefore, the results of vehicle simulation may not contribute sufficiently to the evaluation and problem extraction of odometry sensing algorithms and vehicle control algorithms, and to the improvement of algorithms.

Similarly to the second and third embodiments, the vehicle simulation system of the fourth embodiment also uses a sensor model (which can be called an "odometry sensing statistical model") regarding the movement of the vehicle to simulate the movement of the vehicle. to run. In the vehicle simulation system of the fourth embodiment, (1) an estimated value (can be said to be a quasi-true value close to the actual movement mode of the vehicle) regarding the movement mode of the vehicle calculated using a sensor model (odometry sensing statistical model); , (2) the ideal value related to the mode of movement of the vehicle generated by the vehicle model, and (3) the quasi-ideal value generated by the model based on the information flowing through the in-vehicle network. .

First, the sensor model (odometry sensing statistical model) of the fourth embodiment will be described. The sensor model of the fourth embodiment corresponds to the sensor model of the third embodiment. The sensor model of the fourth embodiment uses, as explanatory variables, information indicating the behavior of the vehicle input to the vehicle when the vehicle is actually traveling, and the difference between the behavior of the vehicle indicated by the information and the actual behavior of the vehicle. is a mathematical model with as the objective variable.

FIG. 13 is a block diagram showing functional blocks of the sensor model generation device 20 of the fourth embodiment that generates the sensor model (odometry sensing statistical model) of the fourth embodiment. The sensor model generation device 20 of the fourth embodiment corresponds to the sensor model generation device 20 of the third embodiment. The sensor model generation device 20 of the fourth embodiment includes a sensor data storage unit 22, a reference data storage unit 24, a tag data storage unit 26, a movement amount derivation unit 120, an indication value derivation unit 130, an error derivation unit 32, an indication value quantum A quantization unit 132 , a tag quantization unit 36 and a model generation unit 38 are provided.

The sensor data storage unit 22 stores, as sensor data, data relating to vehicle behavior detected by a GPS device (also referred to as a position sensor) during an actual test run. The sensor data of the fourth embodiment is data indicating the actual behavior of the vehicle during the test run, and specifically includes a set of vehicle speed and steering angle (which can also be called a steering angle or turning angle). The vehicle may be equipped with an automatic driving controller, and the test driving may be automatic driving controlled by the automatic driving controller.

The reference data storage unit 24 stores the data that flowed through the CAN during the actual test run, that is, the data indicating the behavior of the vehicle that was input to the vehicle during the actual test run as reference data to be compared with the sensor data. Remember. The reference data of the fourth embodiment includes instruction data input from the automatic driving controller to the drive system ECU or the like via CAN during actual test driving, and specifically includes a set of vehicle speed and steering wheel angle.

The tag data storage unit 26 stores, as tag data, various quantitative data or qualitative data that cause at least one of vehicle speed error and steering angle error. The tag data of the fourth embodiment includes the weather (sunny, cloudy, rain, etc.) during the test run, the weight of the vehicle used for the test run, the tire type, and the road surface type. These tag data may be stored in the tag data storage unit 26 by the person in charge of development and testing according to the environment and conditions during test driving.

Hereinafter, "sample" means a data group collected at the same timing during test driving, or data derived from the data group. The movement amount derivation unit 120 derives the movement amount of the vehicle for each sample based on the sensor data stored in the sensor data storage unit 22 . Specifically, the movement amount derivation unit 120 derives the true values of the vehicle speed and the steering angle for each sample. The sensor data storage unit 22 may store sensor data including the amount of lateral movement, the amount of vertical movement of the vehicle per unit time, and the turning angle detected by the GPS device. In this case, the movement amount derivation unit 120 derives the true value of the vehicle speed based on the lateral movement amount and the vertical movement amount indicated by the sensor data, and derives the turning angle indicated by the sensor data as the true value of the steering angle. good too.

Based on the reference data stored in the reference data storage unit 24, the instruction value derivation unit 130 derives an instruction value (an instruction value of the vehicle speed and the steering angle) regarding the movement of the vehicle for each sample. This instruction value corresponds to an ideal value output by the vehicle model unit 18, which will be described later. The reference data storage unit 24 may store reference data including speed instruction data and steering angle instruction data. In this case, the instruction value derivation unit 130 may derive the instruction values of the vehicle speed and the steering angle based on the speed instruction data and the steering angle instruction data indicated by the reference data.

The error derivation unit 32 derives an error between the true value of the vehicle speed derived by the movement amount derivation unit 120 and the command value of the vehicle speed derived by the command value derivation unit 130 . The error deriving section 32 also derives an error between the true value of the steering wheel angle derived by the movement amount deriving section 120 and the indicated value of the steering wheel angle derived by the indicated value deriving section 130 . The error derivation unit 32 derives a set of vehicle speed error and steering wheel angle error for each sample. The error in vehicle speed and the error in the steering angle of the steering wheel are caused by, for example, skidding of tires.

The instruction value quantization unit 132 quantizes the instruction values (the vehicle speed and the steering angle of the steering wheel) regarding the movement of the vehicle for each sample derived by the instruction value derivation unit 130 . Similar to the first to third examples (for example, FIG. 8), the sensor model (odometry sensing statistical model) of the fourth example also includes a plurality of explanatory variables (also called “factors”), and each factor includes a plurality of contains categories. In the fourth embodiment, explanatory variables include vehicle speed and steering angle. The instruction value quantization unit 132 identifies the category to which the vehicle speed and the steering angle correspond to each sample, sets the value of the corresponding category to 1, and sets the value of the non-corresponding category to 0.

The tag quantization section 36 quantizes the tag data for each sample stored in the tag data storage section 26 . Explanatory variables in the sensor model of the fourth embodiment include weather, vehicle weight, tire type, and road surface type. For each sample, the tag quantization unit 36 identifies applicable categories of weather, vehicle weight, tire type, and road surface type, sets the value of the applicable category to 1, and sets the value of the non-applicable category to 0. set.

For each of the plurality of samples, the model generating unit 38 generates quantized data of the command value of the vehicle speed and the command value of the steering angle derived by the command value quantizing unit 132, and the tag derived by the tag quantizing unit 36. A set is generated in which the quantized data of the data is used as an explanatory variable and the vehicle speed error derived by the error derivation unit 32 is used as an objective variable. The model generation unit 38 performs multiple regression analysis based on the plurality of sample data sets to explain the indicated value of vehicle speed, indicated value of steering angle, weather, vehicle weight, tire type, and road surface type. An odometry sensing statistical model is generated as a vehicle speed error model with the vehicle speed error as an objective variable.

In addition, for each of the plurality of samples, the model generation unit 38 generates quantized data of the command value of the vehicle speed and the command value of the steering angle derived by the command value quantization unit 132 and the quantized data of the command value of the steering angle derived by the tag quantization unit 36 A set is generated in which the quantized data of the obtained tag data are used as explanatory variables, and the steering angle error derived by the error deriving section 32 is used as an objective variable. The model generation unit 38 performs multiple regression analysis based on the plurality of sample data sets to explain the indicated value of vehicle speed, indicated value of steering angle, weather, vehicle weight, tire type, and road surface type. A steering angle error model, which is an odometry sensing statistical model with the steering angle error as an objective variable, is generated.

Both the vehicle speed error model and the steering angle error model can be expressed by Equation 1 above. The model generation unit 38 stores the generated vehicle speed error model and steering angle error model in the model storage unit 40 .

Next, a vehicle simulation system using the odometry sensing statistical model generated by the sensor model generation device 20 will be described. FIG. 14 is a block diagram showing functional blocks of the vehicle simulation system 110 of the fourth embodiment. A vehicle simulation system 110 includes a simulation control section 10 , a user interface 12 , a CAN model section 112 , a sensor model section 50 and a vehicle model section 18 .

User interface 12 includes a display device for displaying the results of the simulation. The vehicle model unit 18 derives ideal values regarding the behavior of the vehicle based on the simulation scenario. The vehicle model unit 18 derives the ideal values of the vehicle speed and the steering angle as ideal values regarding the behavior of the vehicle. This ideal value corresponds to the set value of the control command output from the automatic driving controller during automatic driving of the vehicle, and is derived based on the control algorithm of the automatic driving controller.

The CAN model unit 112 is a value corresponding to the ideal value regarding the behavior of the vehicle derived by the vehicle model unit 18 based on the simulation scenario. function as a derivation unit for deriving a quasi-ideal value, which is a value transmitted via the network (CAN in the embodiment). For example, the CAN model unit 112 stores a model or table that associates ideal values (vehicle speed and steering wheel angle) with quasi-ideal values (vehicle speed and steering angle), and uses the model or The table may be referenced to derive quasi-ideal values corresponding to the ideal values.

The quasi-ideal value approximates the ideal value derived by the vehicle model unit 18, but may differ from the ideal value due to limitations on the bit precision of data transmitted by CAN. In addition, the error between the ideal value and the quasi-ideal value accumulates during multiple simulations over multiple unit times (also called “frames”), and the simulation result based on the ideal value and the quasi-ideal value The difference from the simulation results may be large. Note that the vehicle model unit 18 and the CAN model unit 112 may be realized by a known technique (that is, a known vehicle model and CAN model).

The sensor model unit 50 inputs the ideal values relating to the behavior of the vehicle derived by the vehicle model unit 18 to the odometry sensing statistical model generated by the sensor model generation device 20, thereby estimating the behavior of the vehicle in the situation indicated by the simulation scenario. It functions as a second derivation unit that derives estimated values for actual behavior. A detailed configuration of the sensor model unit 50 will be described later.

The simulation control unit 10 controls vehicle simulation. FIG. 15 is a block diagram showing functional blocks of the simulation control unit 10 of the fourth embodiment. The simulation control unit 10 includes a scenario storage unit 140 , a scenario reading unit 142 , an ideal value acquisition unit 144 , a quasi-ideal value acquisition unit 146 , an estimated value acquisition unit 148 , a result image generation unit 150 and an output unit 152 .

A computer program in which the functions of these functional blocks are implemented may be installed in the storage of one or more information processing devices that implement the simulation control unit 10 . Alternatively, the computer program may be installed in the storage of the information processing device via a network. The CPU of the information processing apparatus may display the functions of the functional blocks shown in FIG. 15 by reading the computer program into the main memory and executing it.

The scenario storage unit 140 stores vehicle simulation scenarios. A vehicle simulation scenario is, for example, data indicating vehicle behavior in each of a plurality of unit time "frames", and data indicating weather, vehicle weight, tire type, and road surface type. Scenario reading unit 142 reads the vehicle simulation scenario stored in scenario storage unit 140 for each frame.

The ideal value acquisition unit 144 inputs the vehicle behavior data (in units of frames) indicated by the vehicle simulation scenario read by the scenario reading unit 142 to the vehicle model unit 18 . The ideal value acquisition unit 144 receives the ideal values (in units of frames) of the vehicle speed and the steering angle based on the vehicle simulation scenario output from the vehicle model unit 18 .

The quasi-ideal value acquisition unit 146 inputs to the CAN model unit 112 the ideal values of the vehicle speed and the steering angle (in units of frames) output from the vehicle model unit 18 . The quasi-ideal value acquisition unit 146 receives the quasi-ideal values (in units of frames) of the vehicle speed and steering angle corresponding to the ideal values (in units of frames) of the vehicle speed and the steering angle output from the CAN model unit 112 .

The estimated value acquisition unit 148 inputs the ideal values (in units of frames) of the vehicle speed and steering angle output from the vehicle model unit 18 and the tag data (in units of frames) indicated by the scenario of the vehicle simulation to the sensor model unit 50 . . The estimated value acquisition unit 148 receives the estimated values (frame units) of the vehicle speed and the steering wheel angle corresponding to the ideal values (frame units) of the vehicle speed and the steering wheel angle output from the sensor model unit 50 . The estimated values of the vehicle speed and the steering angle indicate the actual behavior of the vehicle in the situation indicated by the simulation scenario, and can be said to be quasi-true values corresponding to the true values of the actual behavior of the vehicle detected by the GPS device.

The result image generation unit 150 performs a vehicle simulation based on the ideal values output from the vehicle model unit 18, the quasi-ideal values output from the CAN model unit 112, and the estimated values output from the sensor model unit 50. generates a simulation result image, which is an image showing the behavior of the vehicle in An example of the simulation result image will be described later.

The output unit 152 outputs the simulation result image generated by the result image generation unit 150 to the outside. In the fourth embodiment, the output unit 152 causes the display device of the user interface 12 to display the simulation result image. As a modification, the output unit 152 may store the simulation result image in a predetermined storage device.

FIG. 16 is a block diagram showing the details of the sensor model section of the fourth embodiment. The sensor model unit 50 includes a model storage unit 40 , an ideal value quantization unit 160 , a tag quantization unit 52 , an error derivation unit 56 and an estimated value derivation unit 162 . A computer program in which the functions of these functional blocks are implemented may be installed in the storage of one or more information processing devices that implement the sensor model section 50 . Alternatively, the computer program may be installed in the storage of the information processing device via a network. The CPU of the information processing apparatus may display the functions of the functional blocks shown in FIG. 16 by reading the computer program into the main memory and executing it.

The model storage unit 40 stores a vehicle speed error model and a steering angle error model, which are odometry sensing statistical models generated by the sensor model generating device 20 .

The ideal value quantization section 160 corresponds to the indication value quantization section 132 of the sensor model generation device 20 . The ideal value quantization unit 160 quantizes the ideal values (ideal values of vehicle speed and steering angle) regarding the movement of the vehicle for each simulation frame, which are input from the simulation control unit 10 . The ideal value quantization unit 160 identifies, for each frame, a category to which the ideal values of the vehicle speed and steering angle correspond among the plurality of categories defined by the odometry sensing statistical model, and sets the value of the corresponding category to 1. , set the value of non-applicable categories to 0.

A tag quantization unit 52 corresponds to the tag quantization unit 36 of the sensor model generation device 20 . The tag quantization unit 52 quantizes the tag data for each simulation frame input from the simulation control unit 10 . The tag quantization unit 52 identifies, for each frame, the corresponding category of weather, vehicle weight, tire type, and road surface type among a plurality of categories defined by the odometry sensing statistical model, and sets the value of the corresponding category to 1. In addition, the value of the non-applicable category is set to 0.

The error derivation unit 56 converts the quantized data of the ideal values of the vehicle speed and steering angle derived by the ideal value quantization unit 160 and the quantized data of the tag data derived by the tag quantization unit 52 into a vehicle speed error model. to acquire the vehicle speed error output by the vehicle speed error model. Specifically, the error derivation unit 56 reads out the weight of the category for which 1 is set in the quantization data among the plurality of categories of the plurality of explanatory variables defined by the vehicle speed error model, from the model storage unit 40. Using the read weight, the value of the vehicle speed error, which is the objective variable, is obtained. The vehicle speed error is the difference between the speed command value input to the vehicle indicated by the ideal value of the vehicle speed and the actual speed of the vehicle.

Also, the error derivation unit 56 quantizes the quantized data of the ideal values of the vehicle speed and steering angle derived by the ideal value quantization unit 160 and the quantized data of the tag data derived by the tag quantization unit 52 . By inputting to the angle error model, the steering angle error output by the steering angle error model is acquired. Specifically, the error deriving unit 56 reads the weight of the category set to 1 in the quantized data from the multiple categories of the multiple explanatory variables defined by the steering angle error model from the model storage unit 40, Using this read weight, the value of the steering angle error, which is the objective variable, is obtained. The steering angle error is the difference between the command value of the steering wheel angle input to the vehicle indicated by the ideal value of the steering wheel angle and the actual steering angle of the vehicle.

The estimated value derivation unit 162 adds the vehicle speed error derived by the error derivation unit 56 to the ideal value of the vehicle speed for each frame of the simulation input from the simulation control unit 10, thereby obtaining an estimated value of the vehicle speed for each frame of the simulation. to derive In addition, the estimated value derivation unit 162 adds the steering angle error derived by the error derivation unit 56 to the ideal value of the steering wheel steering angle for each frame of the simulation input from the simulation control unit 10 to obtain to derive the estimated value of the steering angle of the steering wheel. The estimated value derivation unit 162 outputs estimated values (that is, estimated values of the vehicle speed and the steering angle of the steering wheel) regarding the movement of the vehicle for each frame of the simulation to the simulation control unit 10 .

In addition, the estimated value derivation unit 162 outputs the estimated value of the vehicle speed for each frame of the simulation and the breakdown information indicating the weight of each of the explanatory variables of the plurality of items in the vehicle speed error model that outputs the vehicle speed error at the time of deriving the estimated value. Output to unit 10 . The breakdown information of the vehicle speed error is the category coefficient of the category used in deriving the vehicle speed error (in other words, the weight for the category set to 1 in the quantized data) among the multiple category coefficients for multiple categories in the vehicle speed error model. (hereinafter referred to as “effective category coefficients”). Specifically, the vehicle speed error breakdown information includes a plurality of sets of explanatory variable item names (for example, “weather”, “vehicle weight”, etc.) and effective category coefficients in the distance error model.

Similarly, the estimated value derivation unit 162 indicates the estimated value of the steering wheel angle for each frame of the simulation and the weight of each explanatory variable of a plurality of items in the steering angle error model that outputs the steering angle error at the time of deriving the estimated value. The breakdown information is output to the simulation control unit 10 . The breakdown information of the steering angle error includes effective category coefficients used in deriving the steering angle error among the plurality of category coefficients for the plurality of categories in the steering angle error model. Specifically, the steering angle error breakdown information includes a plurality of sets of explanatory variable item names (for example, "weather", "vehicle weight", etc.) and effective category coefficients in the steering angle error model.

A simulation result image will be described in detail. FIG. 17 shows an example of a simulation result image. The simulation result image 170 includes a plurality of elements including a vehicle image 172, an ideal value vector 174 (indicated by a dashed line in FIG. 17), a quasi-ideal value vector 176 (indicated by a dashed line in FIG. 17), and an estimated value vector 178 (indicated by a dashed line in FIG. 17). shown by a solid line).

The result image generation unit 150 of the simulation control unit 10 sets an ideal value vector 174 indicating the ideal values of the behavior of the vehicle in the simulation result image 170 based on the ideal values of the vehicle speed and steering angle for each frame of the vehicle simulation. do. In addition, the result image generation unit 150 sets a quasi-ideal value vector 176 indicating quasi-ideal values of the behavior of the vehicle in the simulation result image 170 based on the quasi-ideal values of the vehicle speed and the steering angle for each frame of the vehicle simulation. do. In addition, the result image generation unit 150 generates an estimated value vector 178 representing an estimated value (in other words, a quasi-true value) of the behavior of the vehicle based on the estimated values of the vehicle speed and the steering angle for each frame of the vehicle simulation. Set to image 170 .

The result image generation unit 150 sets the ideal value vector 174, the quasi-ideal value vector 176, and the estimated value vector 178 in different modes (styles). For example, the ideal value vector 174, the quasi-ideal value vector 176, and the estimated value vector 178 are set so that at least one of line type, shape, pattern, color, thickness, presence/absence of shading, and the like is different from each other. As a result, it is possible to provide a simulation result image in which the ideal value, the quasi-ideal value, and the estimated value of the behavior of the vehicle can be intuitively distinguished, and the difference between them can be easily grasped.

In addition, the result image generator 150 sets an element indicating the difference between the ideal value and the estimated value of the behavior of the vehicle in the simulation result image 170 . Specifically, the result image generator 150 sets a steering angle error indicator indicating the difference in steering angle (steering angle error) and a vehicle speed error indicator indicating the difference in vehicle speed (vehicle speed error) in the simulation result image 170. do. The simulation result image 170 of FIG. 17 shows the steering angle error indicator 180 .

Although not shown in FIG. 17, the result image generation unit 150 may set in the simulation result image 170 an element that indicates the difference between the quasi-ideal value and the estimated value of the behavior of the vehicle. Specifically, result image generator 150 may set a steering angle error indicator and a vehicle speed error indicator regarding the difference between the quasi-ideal value and the estimated value in simulation result image 170 . By arranging the steering angle error indicator and the vehicle speed error indicator in this way, it is possible to easily grasp the steering angle error and the vehicle speed error in the simulation results.

FIG. 18 also shows an example of a simulation result image. As shown in FIG. 18, the simulation result image 170 is configured to display the details of the difference between the ideal value and the estimated value of the behavior of the vehicle according to a predetermined operation.

For example, the result image generator 150 simulates the error breakdown 182, which is an image showing the breakdown of the steering angle error, when the steering angle error indicator 180 is selected (eg, moused over) on the simulation result image 170. A JavaScript (registered trademark) code to be additionally displayed on the result image 170 is set in the data of the simulation result image 170 . Further, although not shown in FIG. 18, when the vehicle speed error indicator is selected in the simulation result image 170, the result image generation unit 150 displays the error breakdown, which is an image showing the breakdown of the vehicle speed error, on the simulation result image 170. A JavaScript code to be additionally displayed is set in the data of the simulation result image 170 .

As described above, the estimated value derivation unit 162 of the sensor model unit 50 not only estimates the vehicle speed and the steering angle for each frame, but also provides the breakdown information of the vehicle speed error and the steering angle error for each frame to the simulation control unit 10. Output to The breakdown information includes a plurality of pairs of explanatory variable item names and effective category coefficients in the vehicle speed error model or the steering angle error model.

As shown in FIG. 18, the result image generation unit 150 simulates an error breakdown 182 including a pie chart indicating the ratio of effective category coefficients of a plurality of explanatory variables in the steering angle error model in association with the steering angle error indicator 180. Set to the data of the result image 170 . The error breakdown 182 in FIG. 18 shows weather, vehicle weight, tire type, and road surface type as explanatory variable items, but other items may be included. Appropriate items may be selected through experiments or the like using the simulation system 110 .

Further, although not shown in FIG. 18, the result image generation unit 150 associates the error breakdown including the pie chart showing the ratio of the effective category coefficients of the multiple explanatory variables in the vehicle speed error model with the vehicle speed error indicator, and the simulation result Set to the data of image 170 . The error breakdown (for example, the error breakdown 182) displayed in the simulation result image 170 can be said to be a statistical chart showing the weights (in other words, the degree of relative influence) of each of the multiple explanatory variables in the steering angle error model or vehicle speed error model.

FIG. 19 also shows an example of a simulation result image. Although the error breakdown 182 of FIG. 18 shows the ratio of effective category coefficients for each explanatory variable as a pie chart, the error breakdown may include a different type of statistical chart than a pie chart. The error breakdown 184 in FIG. 19 shows the ratio of the effective category coefficients of the multiple explanatory variables in the steering angle error model and the ratio of the effective category coefficients of the multiple explanatory variables in the vehicle speed error model by stacked bar graphs. In the simulation result image 170 of FIG. 19, the same error breakdown 184 may be displayed regardless of whether the steering angle error indicator 180 or the vehicle speed error indicator 181 is selected.

The error breakdown may indicate information other than the ratio of the effective category coefficients of each explanatory variable of the steering angle error model or vehicle speed error model. For example, the error breakdown may indicate only the name of each explanatory variable of the steering angle error model or the vehicle speed error model, or may indicate the value of the steering angle error or the vehicle speed error.

FIG. 20 also shows an example of a simulation result image. The simulation result image 170 in FIG. 20 is the difference between the ideal value and the estimated value of the behavior of the vehicle ( ie accumulated error).

The result image generation unit 150 sets an ideal value synthetic vector 175 obtained by synthesizing a plurality of ideal value vectors 174 in the simulation result image 170 . The ideal value synthetic vector 175 in FIG. 20 is obtained by synthesizing the ideal value vectors 174 for three frames. In addition, the result image generator 150 sets an estimated value combined vector 179 obtained by synthesizing a plurality of estimated value vectors 178 in the simulation result image 170 . The estimated value composite vector 179 in FIG. 20 is obtained by synthesizing the estimated value vectors 178 for three frames. Furthermore, the result image generator 150 sets a steering angle error indicator 180 and a vehicle speed error indicator 181 indicating the difference (that is, accumulated error) between the ideal value combined vector 175 and the estimated value combined vector 179 in the simulation result image 170 .

As in FIG. 19 , the result image generator 150 generates a simulation result image such that the error breakdown 184 is popped up when the steering angle error indicator 180 is selected or the vehicle speed error indicator 181 is selected. 170 is set. Specifically, the result image generation unit 150 may obtain the average value of the effective category coefficients for the synthesized frames based on the effective category coefficients of the multiple explanatory variables in the steering angle error model. The result image generation unit 150 generates an error breakdown 184 of the steering angle error that indicates the ratio of the effective category coefficients (average values) of the multiple explanatory variables in the steering angle error model, and associates the steering angle error indicator 180 with the simulation result. Set to the data of image 170 .

Further, the result image generator 150 may obtain the average value of the effective category coefficients for frames to be synthesized based on the effective category coefficients of the multiple explanatory variables in the vehicle speed error model. The result image generation unit 150 generates an error breakdown 184 of the vehicle speed error that indicates the ratio of the effective category coefficients (average values) of the multiple explanatory variables in the vehicle speed error model, and associates it with the vehicle speed error indicator 181 to generate the simulation result image 170. Set to data.

Note that the simulation result image 170 may include an element indicating the difference between the quasi-ideal vector 176 and the estimated value vector 178 . Also, the simulation result image 170 may be configured to display the error breakdown between the quasi-ideal value and the estimated value when the element is selected. The details of the error between the quasi-ideal value and the estimated value in a certain frame may be the same as the details of the error between the ideal value and the estimated value in the same frame. Furthermore, the simulation results image 170 may be configured to show the difference between the quasi-ideal and estimated values accumulated over multiple frames of the vehicle simulation.

The present disclosure has been described above based on the fourth embodiment. The fourth embodiment is an example, and it should be understood by those skilled in the art that various modifications can be made to the combination of each component or each treatment process, and such modifications are within the scope of the present disclosure.

Although ideal values, quasi-ideal values, and estimated values are displayed simultaneously in the simulation result image 170 of the fourth embodiment, as a modification, the types of values displayed in the simulation result image 170 can be selected by the user. good. For example, the result image generation unit 150 of the simulation control unit 10 may generate the simulation result image 170 including the ideal values and the estimated values but not including the quasi-ideal values according to the user's operation. Further, the result image generation unit 150 may generate the simulation result image 170 that includes the quasi-ideal values and the estimated values but does not include the ideal values according to the user's operation.

Further, although not mentioned in the fourth embodiment, the orientation of the vehicle image 172 in the simulation result image 170, in other words, the traveling direction of the vehicle image 172 in the simulation result image 170 may be changeable by the user. . For example, when the default direction of the vehicle image 172 in the simulation result image 170 is rightward, the result image generation unit 150 changes the direction of the vehicle image 172 in the simulation result image 170 to upward or the like according to the user's operation. may This can assist analysis of simulation results.

Of course, the curve fitting described in the first modification of the first embodiment and the addition of white noise described in the second modification of the first embodiment can also be applied as modifications of the fourth embodiment. be.

The techniques described in the fourth embodiment and modifications may be specified by the following items.
[Item 2-1]
A model in which information indicating the behavior of the vehicle input to the vehicle during actual driving is used as an explanatory variable, and the difference between the behavior of the vehicle indicated by the information and the actual behavior of the vehicle is used as the objective variable. a storage unit that stores
a derivation unit that derives an estimated value regarding the actual behavior of the vehicle in a situation indicated by the simulation scenario by inputting into the model ideal values regarding the behavior of the vehicle derived based on the simulation scenario;
a generator that generates an image showing the ideal value and the estimated value in different modes as an image showing the behavior of the vehicle in the simulation;
vehicle simulation system.
According to this vehicle simulation system, in the image of the simulation result, the ideal value of the behavior of the vehicle (for example, the indicated value in automatic driving of the vehicle) and the estimated value of the behavior of the vehicle (for example, the value indicating the actual behavior of the vehicle). can be easily distinguished, it is possible to provide useful simulation results for evaluating odometry sensing algorithms and vehicle control algorithms, extracting issues, and improving the algorithms.
[Item 2-2]
the image includes elements that indicate the difference between the ideal value and the estimated value;
A vehicle simulation system according to item 2-1.
According to this vehicle simulation system, it becomes easier for the user to intuitively grasp the difference between the ideal value and the estimated value of the behavior of the vehicle.
[Item 2-3]
wherein the image is configured to display a breakdown of the difference between the ideal value and the estimated value in response to a predetermined operation;
A vehicle simulation system according to item 2-1 or 2-2.
According to this vehicle simulation system, it is possible to support the analysis of the difference between the ideal value and the estimated value of the behavior of the vehicle.
[Item 2-4]
The model defines weights for each of a plurality of explanatory variables,
The breakdown of the difference between the ideal value and the estimated value indicates the weight of each of the plurality of explanatory variables,
A vehicle simulation system according to item 2-3.
According to this vehicle simulation system, it is possible to effectively analyze the difference between the ideal and estimated values by showing multiple factors that caused the difference between the ideal and estimated values of vehicle behavior and the degree of influence of each factor. can support
[Item 2-5]
The image shows the difference between the ideal value and the estimated value accumulated over multiple frames of the simulation.
A simulation system according to any one of items 2-1 to 2-4.
According to this vehicle simulation system, by showing the cumulative error between the ideal and estimated values of the behavior of the vehicle over a plurality of frames, the difference between the ideal and estimated values becomes clearer. It can effectively support the analysis on the difference between the estimated value and the
[Item 2-6]
A model in which information indicating the behavior of the vehicle input to the vehicle during actual driving is used as an explanatory variable, and the difference between the behavior of the vehicle indicated by the information and the actual behavior of the vehicle is used as the objective variable. a storage unit that stores
A value corresponding to an ideal value regarding the behavior of the vehicle derived based on a simulation scenario, which is transmitted via a network within the vehicle in order to cause the vehicle to execute the behavior indicated by the ideal value. a first derivation unit that derives a quasi-ideal value that is a value;
a second derivation unit that derives an estimated value regarding the actual behavior of the vehicle in a situation indicated by the scenario of the simulation by inputting ideal values regarding the behavior of the vehicle into the model;
a generation unit that generates an image showing the quasi-ideal value and the estimated value in different modes as an image showing the behavior of the vehicle in the simulation;
vehicle simulation system.
According to this vehicle simulation system, in the image of the simulation result, the semi-ideal value of the behavior of the vehicle (for example, the value input to each ECU via CAN in automatic operation of the vehicle) and the estimated value of the behavior of the vehicle (for example, values that indicate the actual behavior of the vehicle), it is possible to provide useful simulation results for evaluating odometry sensing algorithms and vehicle control algorithms, extracting problems, and improving the algorithms.
[Item 2-7]
A model stored in a storage unit, in which information indicating the behavior of the vehicle input to the vehicle when the vehicle is actually traveling is used as an explanatory variable, and the behavior of the vehicle indicated by the information and the actual behavior of the vehicle are used. By inputting the ideal values regarding the behavior of the vehicle derived based on the simulation scenario into a model whose objective variable is the difference between the behavior of the vehicle and the actual behavior of the vehicle in the situation indicated by the simulation scenario derive an estimate for
generating an image showing the ideal value and the estimated value in a different manner as an image showing the behavior of the vehicle in the simulation;
A computer-implemented vehicle simulation method.
This vehicle simulation method has the same effect as the vehicle simulation system of item 2-1.
[Item 2-8]
A value corresponding to an ideal value regarding vehicle behavior derived based on a simulation scenario, which value is transmitted via a network within the vehicle in order to cause the vehicle to execute the behavior indicated by the ideal value. Derive a quasi-ideal value that is
A model stored in a storage unit, in which information indicating the behavior of the vehicle input to the vehicle when the vehicle is actually traveling is used as an explanatory variable, and the behavior of the vehicle indicated by the information and the actual behavior of the vehicle are used. By inputting an ideal value for the behavior of the vehicle into a model whose objective variable is the difference between the behavior of
generating an image showing the quasi-ideal value and the estimated value in different manners as an image showing the behavior of the vehicle in the simulation;
A computer-implemented vehicle simulation method.
This vehicle simulation method has the same effect as the vehicle simulation system of item 2-6.
[Item 2-9]
A model stored in a storage unit, in which information indicating the behavior of the vehicle input to the vehicle when the vehicle is actually traveling is used as an explanatory variable, and the behavior of the vehicle indicated by the information and the actual behavior of the vehicle are used. By inputting the ideal values regarding the behavior of the vehicle derived based on the simulation scenario into a model whose objective variable is the difference between the behavior of the vehicle and the actual behavior of the vehicle in the situation indicated by the simulation scenario derive an estimate for
generating an image showing the ideal value and the estimated value in different manners as an image showing the behavior of the vehicle in the simulation;
A computer program that makes a computer do something.
This computer program has the same effects as the vehicle simulation system of item 2-1.
[Item 2-10]
A value corresponding to an ideal value regarding the behavior of a vehicle derived based on a simulation scenario, which value is transmitted via a network within the vehicle in order to cause the vehicle to execute the behavior indicated by the ideal value. Derive a quasi-ideal value that is
A model stored in a storage unit, in which information indicating the behavior of the vehicle that is input to the vehicle when the vehicle is actually traveling is used as an explanatory variable, and the behavior of the vehicle indicated by the information and the actual behavior of the vehicle are used. By inputting ideal values regarding the behavior of the vehicle into a model whose objective variable is the difference between the behavior of
generating an image showing the quasi-ideal value and the estimated value in different manners as an image showing the behavior of the vehicle in the simulation;
A computer program that makes a computer do something.
This computer program has the same effect as the vehicle simulation system of item 2-6.

Any combination of the above-described examples and modifications is also useful as an embodiment of the present disclosure. A new embodiment resulting from the combination has the effects of each combined embodiment and modified example. It should also be understood by those skilled in the art that the functions to be fulfilled by each constituent element described in the claims are realized by each constituent element shown in the embodiments and modified examples singly or in conjunction with each other.

12 user interface 18 vehicle model section 20 sensor model generation device 40 model storage section 50 sensor model section 110 vehicle simulation system 112 CAN model section 150 result image generation section.

Claims (10)

A model in which information indicating the behavior of the vehicle input to the vehicle during actual driving is used as an explanatory variable, and the difference between the behavior of the vehicle indicated by the information and the actual behavior of the vehicle is used as the objective variable. a storage unit that stores
a derivation unit that derives an estimated value regarding the actual behavior of the vehicle in a situation indicated by the simulation scenario by inputting into the model ideal values regarding the behavior of the vehicle derived based on the simulation scenario;
a generator that generates an image showing the ideal value and the estimated value in different modes as an image showing the behavior of the vehicle in the simulation;
vehicle simulation system. the image includes elements that indicate the difference between the ideal value and the estimated value;
The vehicle simulation system according to claim 1. wherein the image is configured to display a breakdown of the difference between the ideal value and the estimated value in response to a predetermined operation;
A vehicle simulation system according to claim 1 or 2. The model defines weights for each of a plurality of explanatory variables,
The breakdown of the difference between the ideal value and the estimated value indicates the weight of each of the plurality of explanatory variables,
A vehicle simulation system according to claim 3. The image shows the difference between the ideal value and the estimated value accumulated over multiple frames of the simulation.
A vehicle simulation system according to any one of claims 1 to 4. A model in which information indicating the behavior of the vehicle input to the vehicle during actual driving is used as an explanatory variable, and the difference between the behavior of the vehicle indicated by the information and the actual behavior of the vehicle is used as the objective variable. a storage unit that stores
A value corresponding to an ideal value regarding the behavior of the vehicle derived based on a simulation scenario, which is transmitted via a network within the vehicle in order to cause the vehicle to execute the behavior indicated by the ideal value. a first derivation unit that derives a quasi-ideal value that is a value;
a second derivation unit that derives an estimated value regarding the actual behavior of the vehicle in a situation indicated by the scenario of the simulation by inputting ideal values regarding the behavior of the vehicle into the model;
a generation unit that generates an image showing the quasi-ideal value and the estimated value in different modes as an image showing the behavior of the vehicle in the simulation;
vehicle simulation system. A model stored in a storage unit, in which information indicating the behavior of the vehicle input to the vehicle when the vehicle is actually traveling is used as an explanatory variable, and the behavior of the vehicle indicated by the information and the actual behavior of the vehicle are used. By inputting the ideal values regarding the behavior of the vehicle derived based on the simulation scenario into a model whose objective variable is the difference between the behavior of the vehicle and the actual behavior of the vehicle in the situation indicated by the simulation scenario derive an estimate for
generating an image showing the ideal value and the estimated value in a different manner as an image showing the behavior of the vehicle in the simulation;
A computer-implemented vehicle simulation method. A value corresponding to an ideal value regarding vehicle behavior derived based on a simulation scenario, which value is transmitted via a network within the vehicle in order to cause the vehicle to execute the behavior indicated by the ideal value. Derive a quasi-ideal value that is
A model stored in a storage unit, in which information indicating the behavior of the vehicle input to the vehicle when the vehicle is actually traveling is used as an explanatory variable, and the behavior of the vehicle indicated by the information and the actual behavior of the vehicle are used. By inputting an ideal value for the behavior of the vehicle into a model whose objective variable is the difference between the behavior of
generating an image showing the quasi-ideal value and the estimated value in different manners as an image showing the behavior of the vehicle in the simulation;
A computer-implemented vehicle simulation method. A model stored in a storage unit, in which information indicating the behavior of the vehicle input to the vehicle when the vehicle is actually traveling is used as an explanatory variable, and the behavior of the vehicle indicated by the information and the actual behavior of the vehicle are used. By inputting the ideal values regarding the behavior of the vehicle derived based on the simulation scenario into a model whose objective variable is the difference between the behavior of the vehicle and the actual behavior of the vehicle in the situation indicated by the simulation scenario derive an estimate for
generating an image showing the ideal value and the estimated value in a different manner as an image showing the behavior of the vehicle in the simulation;
A computer program that makes a computer do something. A value corresponding to an ideal value regarding vehicle behavior derived based on a simulation scenario, which value is transmitted via a network within the vehicle in order to cause the vehicle to execute the behavior indicated by the ideal value. Derive a quasi-ideal value that is
A model stored in a storage unit, in which information indicating the behavior of the vehicle input to the vehicle when the vehicle is actually traveling is used as an explanatory variable, and the behavior of the vehicle indicated by the information and the actual behavior of the vehicle are used. By inputting an ideal value for the behavior of the vehicle into a model whose objective variable is the difference between the behavior of
generating an image showing the quasi-ideal value and the estimated value in different manners as an image showing the behavior of the vehicle in the simulation;
A computer program that makes a computer do something.

Images