An optimization method, device, medium and equipment of a vehicle collision simulation model

Abstract

The application provides a vehicle collision simulation model optimization method, device, medium and equipment, an initial model of a simulation vehicle is constructed based on vehicle parameters of the simulation vehicle; the initial model is preliminarily trained to obtain a preliminary training model; the preliminary training model is further trained to obtain a target model of the simulation vehicle; that is, the initial model of the simulation vehicle is constructed, and the initial model is preliminarily trained by using historical collision simulation sequence data, so that the simulation data with a large amount of data is used for rapid training, the vehicle model quickly reaches a certain precision, and the preliminary training model is further trained by using test sequence data, so that the test data with a small amount of data is used to improve the precision of the vehicle model, and the historical collision simulation sequence data with a large correlation with the test sequence data is selected to improve the training precision of the initial model, thereby improving the efficiency and effect of model training and optimization.

Description

Technical Field

[0001] This application relates to the field of automotive collision simulation technology, specifically to an optimization method, apparatus, medium, and equipment for an automotive collision simulation model. Background Technology

[0002] The simulation of vehicle collision scenarios is mainly achieved through computer simulation software. By simulating data such as the deformation of the vehicle body structure during a collision, the structure and constraint system are analyzed and optimized, thereby improving vehicle safety and reducing the injury value of occupants.

[0003] Currently, the finite element method is mainly used for vehicle structural safety simulation both domestically and internationally. The more meshes used in the finite element method, the higher the accuracy of the simulation calculation. Therefore, the mesh size of the vehicle models built by automotive engineers is often large, resulting in a long calculation time. Furthermore, the benchmarking of the vehicle structure requires a large number of case studies, which leads to an excessive amount of computation for the benchmarking of the vehicle structural safety simulation model, prolonging the construction time of the vehicle structural simulation model and reducing the efficiency of vehicle collision simulation. Summary of the Invention

[0004] To address the aforementioned technical problems, this application is proposed. Embodiments of this application provide a method, apparatus, medium, and device for optimizing an automobile collision simulation model.

[0005] According to one aspect of this application, an optimization method for a car crash simulation model is provided, comprising: constructing an initial model of the simulated vehicle based on vehicle parameters of the simulated vehicle; inputting historical crash simulation sequence data into the initial model to perform preliminary training on the initial model to obtain a pre-trained model; inputting experimental test sequence data into the pre-trained model to perform retraining on the pre-trained model to obtain a target model of the simulated vehicle; wherein the correlation between the historical crash simulation sequence data and the experimental test sequence data is greater than a preset correlation threshold.

[0006] In one embodiment, the step of inputting historical collision simulation sequence data into the initial model to perform preliminary training on the initial model to obtain a pre-trained model includes: inputting historical collision simulation sequence data into the initial model to obtain a first output result; calculating a first loss value based on the first output result and the observation value corresponding to the historical collision simulation sequence data; if the first loss value is greater than a preset first loss value threshold, adjusting the parameters of the initial model until the first loss value is less than or equal to the first loss value threshold to obtain the pre-trained model.

[0007] In one embodiment, adjusting the parameters of the initial model includes: calculating the objective function values ​​of multiple data points in the historical collision simulation sequence data; wherein the objective function represents the first loss value of the corresponding data; calculating the optimal objective function value, the mean of the objective function value, and the standard deviation of the objective function value of the historical collision simulation sequence data; calculating the parameter adjustment step size based on the optimal objective function value, the mean of the objective function value, and the standard deviation of the objective function value; and adjusting the parameters of the initial model based on the parameter adjustment step size.

[0008] In one embodiment, calculating the parameter adjustment step size based on the optimal objective function value, the mean of the objective function value, and the standard deviation of the objective function value includes: the calculation formula for the parameter adjustment step size is:

[0009] ;

[0010] in, Adjust the step size for the parameters. For regularization functions, The mean of the objective function values. To obtain the optimal objective function value, The standard deviation of the objective function value. The threshold value is set.

[0011] In one embodiment, the historical collision simulation sequence data includes multiple data sets; wherein, the step of inputting the historical collision simulation sequence data into the initial model to perform preliminary training on the initial model to obtain the initial training model includes: calculating the correlation between each historical collision simulation sequence data set and the experimental test sequence data; if the correlation is greater than the correlation threshold, selecting the corresponding historical collision simulation sequence data set as the target collision simulation sequence data set; and inputting the target collision simulation sequence data into the initial model to perform preliminary training on the initial model to obtain the initial training model.

[0012] In one embodiment, the step of inputting the test sequence data into the initial training model to retrain the initial training model and obtain the target model of the simulated vehicle includes: inputting the test sequence data into the initial training model to obtain a second output result; calculating a second loss value based on the second output result and the observation value corresponding to the test sequence data; if the second loss value is greater than a preset second loss value threshold, adjusting the parameters of the initial training model until the second loss value is less than or equal to the second loss value threshold to obtain the target model.

[0013] In one embodiment, adjusting the parameters of the initial training model includes: adjusting the parameter values ​​of the initial training model; calculating the parameter error variable of the initial training model; and calculating the parameter error of the target model based on the parameter error of the initial training model and the parameter error variable.

[0014] According to another aspect of this application, an optimization apparatus for a car collision simulation model is provided, comprising: a model building module for building an initial model of the simulation vehicle based on vehicle parameters of the simulation vehicle; a preliminary training module for inputting historical collision simulation sequence data into the initial model to perform preliminary training on the initial model to obtain a pre-trained model; and a retraining module for inputting experimental test sequence data into the pre-trained model to perform retraining on the pre-trained model to obtain a target model of the simulation vehicle; wherein the correlation between the historical collision simulation sequence data and the experimental test sequence data is greater than a preset correlation threshold.

[0015] According to another aspect of this application, a computer-readable storage medium is provided, the storage medium storing a computer program for performing any of the methods described above.

[0016] According to another aspect of this application, an electronic device is provided, comprising: a processor; a memory for storing processor-executable instructions; the processor being configured to perform any of the methods described above.

[0017] This application provides a method, apparatus, medium, and equipment for optimizing a car collision simulation model. It constructs an initial model of the simulated vehicle based on the vehicle parameters; inputs historical collision simulation sequence data into the initial model for preliminary training, obtaining a pre-trained model; and inputs experimental test sequence data into the pre-trained model for further training, obtaining a target model of the simulated vehicle. The correlation between the historical collision simulation sequence data and the experimental test sequence data is greater than a preset correlation threshold. Specifically, by constructing an initial model of the simulated vehicle and using historical collision simulation sequence data for preliminary training (using a large amount of simulation data for rapid training) to quickly achieve a certain level of accuracy, and further training the pre-trained model using experimental test sequence data (using a smaller amount of test data to improve the accuracy of the vehicle model), and by selecting historical collision simulation sequence data with a high correlation to the experimental test sequence data, the training accuracy of the initial model is improved, thereby enhancing the efficiency and effectiveness of model training and optimization. Attached Figure Description

[0018] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0019] Figure 1 This is a flowchart illustrating an exemplary embodiment of the method for optimizing a car collision simulation model provided in this application.

[0020] Figure 2 This is a schematic diagram of the structure of an optimization device for a car collision simulation model provided in an exemplary embodiment of this application.

[0021] Figure 3 This is a structural diagram of an electronic device provided in an exemplary embodiment of this application. Detailed Implementation

[0022] Hereinafter, exemplary embodiments according to this application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments of this application. It should be understood that this application is not limited to the exemplary embodiments described herein.

[0023] Figure 1 This is a flowchart illustrating an exemplary embodiment of the method for optimizing a car collision simulation model provided in this application. Figure 1 As shown, the optimization method for this car crash simulation model includes the following steps:

[0024] Step 110: Based on the vehicle parameters of the simulated vehicle, construct the initial model of the simulated vehicle.

[0025] This application is based on finite element software, which combines vehicle parameters of the simulated vehicle (such as vehicle type, shape, size and stiffness) to establish a whole vehicle model according to the collision conditions. The software includes, but is not limited to, Perimer, HyperWorks, ANSA, etc. The collision conditions include, but are not limited to, frontal 50 km / h 100% collision (50FRB), frontal 50 km / h 40% overlap deformable barrier collision (64ODB), frontal 50% overlap moving progressive deformable barrier collision test (50MPDB), etc.

[0026] Step 120: Input the historical collision simulation sequence data into the initial model to perform preliminary training on the initial model and obtain the initial trained model.

[0027] After constructing an initial model of the simulated vehicle, this application acquires historical collision simulation sequence data and inputs this data into the initial model for preliminary training, resulting in a pre-trained model. The training process may involve parameterizing the initial model to obtain a model function.

[0028]

[0029] in, To input the observed numerical function, To pass through variables x The function equations of the initial model ( x These can be controllable or uncontrollable variables. Controllable variables include the product's size or shape, while uncontrollable variables include material properties, structural quality, and stiffness. For unknown regression coefficients, This refers to model bias or model error. It is the measurement error (assuming it is Gaussian noise, and...) x (Irrelevant)

[0030] This application can define =1 is used to simplify the training process of the simulation model. The simplified model function is: This application controls errors. To improve the accuracy of the simulation model.

[0031] Step 130: Input the test sequence data into the initial training model to retrain the initial training model and obtain the target model of the simulated vehicle.

[0032] In this application, the correlation between historical collision simulation sequence data and experimental test sequence data is greater than a preset correlation threshold. After initially training the initial model using historical collision simulation sequence data, the application further trains the initially trained model using experimental test sequence data whose correlation with the historical collision simulation sequence data is greater than a preset correlation threshold, thereby improving the accuracy of the simulation model and obtaining a high-precision simulation model.

[0033] This application provides an optimization method for a car collision simulation model. The method involves constructing an initial model of the simulated vehicle based on its parameters; inputting historical collision simulation sequence data into the initial model for preliminary training to obtain a pre-trained model; and then inputting experimental test sequence data into the pre-trained model for further training to obtain a target model of the simulated vehicle. The correlation between the historical collision simulation sequence data and the experimental test sequence data is greater than a preset correlation threshold. Specifically, by constructing an initial model of the simulated vehicle and using historical collision simulation sequence data for preliminary training (using a large amount of simulation data for rapid training) to quickly achieve a certain level of accuracy, and further training the pre-trained model using experimental test sequence data (using a smaller amount of test data to improve the accuracy of the vehicle model), and by selecting historical collision simulation sequence data with a high correlation to the experimental test sequence data, the training accuracy of the initial model is improved, thereby enhancing the efficiency and effectiveness of model training and optimization.

[0034] In one embodiment, the specific implementation of step 120 above may be as follows: inputting historical collision simulation sequence data into the initial model to obtain a first output result; calculating a first loss value based on the first output result and the observation value corresponding to the historical collision simulation sequence data; if the first loss value is greater than a preset first loss value threshold, adjusting the parameters of the initial model until the first loss value is less than or equal to the first loss value threshold to obtain the initial training model.

[0035] This application uses the initial parameters of an initial model to predict historical collision simulation sequence data to obtain a first output result. Then, it calculates the difference between the first output result and the corresponding observed values ​​of the historical collision simulation sequence data to obtain a first loss value. If the first loss value is large, the parameters of the initial model are adjusted until the first output result and the corresponding observed values ​​of the historical collision simulation sequence data are close, indicating that the initial model training is complete. The formula for calculating the first loss value (i.e., the objective function) is as follows:

[0036]

[0037] in, J The first loss value, For the first The observations corresponding to the historical collision simulation sequence data. This is the first output result. The error parameter to be adjusted (corresponding to the error mentioned above). This represents the number of historical collision simulation sequence data.

[0038] In one embodiment, step 120 can be implemented as follows: calculating the objective function values ​​of multiple data in the historical collision simulation sequence data; wherein the objective function represents the first loss value of the corresponding data; calculating the optimal objective function value, the mean of the objective function value, and the standard deviation of the objective function value of the historical collision simulation sequence data; calculating the parameter adjustment step size based on the optimal objective function value, the mean of the objective function value, and the standard deviation of the objective function value; and adjusting the parameters of the initial model based on the parameter adjustment step size.

[0039] This application calculates the objective function values ​​of multiple data points in historical collision simulation sequence data, and then calculates the optimal objective function value, mean objective function value, and standard deviation of the objective function value based on the objective function values. Combining the optimal objective function value, mean objective function value, and standard deviation of the objective function value, the parameter adjustment step size is calculated. The parameters of the initial model are adjusted according to the parameter adjustment step size, thereby realizing iterative updating of the parameters of the initial model.

[0040] In one embodiment, step 120 can be specifically implemented as follows: the formula for calculating the parameter adjustment step size is:

[0041] ;

[0042] in, Adjust the step size for the parameters. For regularization functions, The mean of the objective function values. To obtain the optimal objective function value, The standard deviation of the objective function value. The threshold value is set.

[0043] In one embodiment, the historical collision simulation sequence data includes multiple data sets; wherein, the specific implementation of step 120 above may be: calculating the correlation between each historical collision simulation sequence data set and the experimental test sequence data; if the correlation is greater than the correlation threshold, then selecting the corresponding historical collision simulation sequence data set as the target collision simulation sequence data set; inputting the target collision simulation sequence data set into the initial model to perform preliminary training on the initial model, thereby obtaining the initial training model.

[0044] After acquiring multiple historical collision simulation sequence data, this application, recognizing that different historical collision simulation sequence data may correspond to different vehicle models and parameters, filters the acquired historical collision simulation sequence data to obtain the historical collision simulation sequence data most closely resembling the simulated vehicle. Specifically, this application calculates the correlation between historical collision simulation sequence data and experimental test sequence data. If the correlation is greater than a correlation threshold, the corresponding historical collision simulation sequence data is selected as the target collision simulation sequence data. All the selected target collision simulation sequence data are then input into the initial model for preliminary training, resulting in the initial trained model. The formula for calculating the correlation between historical collision simulation sequence data and experimental test sequence data is as follows:

[0045] ;

[0046] in, For correlation, and These are the first two sets of historical collision simulation sequence data and experimental test sequence data, respectively. i One value, and These are the average values ​​from historical collision simulation sequence data and experimental test sequence data, respectively. This represents the number of historical collision simulation sequence data.

[0047] In one embodiment, the specific implementation of step 130 above may be as follows: inputting experimental test sequence data into the initial training model to obtain a second output result; calculating a second loss value based on the second output result and the observation value corresponding to the experimental test sequence data; if the second loss value is greater than a preset second loss value threshold, adjusting the parameters of the initial training model until the second loss value is less than or equal to the second loss value threshold to obtain the target model.

[0048] This application uses an initial training model to predict historical collision simulation sequence data to obtain a second output result. Then, it calculates the difference between the second output result and the observation value corresponding to the experimental test sequence data to obtain a second loss value. If the second loss value is large, the parameters of the initial training model are adjusted until the second output result and the observation value corresponding to the experimental test sequence data are close, which indicates that the initial training model has been trained.

[0049] In one embodiment, step 130 can be implemented by: adjusting the parameter values ​​of the initial training model; calculating the parameter error variable of the initial training model; and calculating the parameter error of the target model based on the parameter error and parameter error variable of the initial training model.

[0050] In this application, when adjusting the initial training model, the parameters of the initial training model are divided into parameter values ​​and parameter errors. The adjustment of the initial training model is achieved by adjusting the parameter values ​​and parameter errors. Specifically, after the initial training model is obtained, when the initial training model is retrained using experimental test sequence data, the parameter error of the initial training model is fixed, and the parameter error of the final target model is modified by the additional error generated in the experimental test sequence data (i.e., parameter error variable). That is, the parameter error of the target model = the parameter error of the initial training model + the parameter error variable.

[0051] Figure 2 This is a schematic diagram of the structure of an optimization device for a car collision simulation model provided in an exemplary embodiment of this application. Figure 2 As shown, the optimization device 20 for the car collision simulation model includes: a model building module 21, used to build an initial model of the simulation vehicle based on the vehicle parameters of the simulation vehicle; a preliminary training module 22, used to input historical collision simulation sequence data into the initial model to perform preliminary training on the initial model to obtain a preliminary training model; and a retraining module 23, used to input experimental test sequence data into the preliminary training model to perform retraining on the preliminary training model to obtain a target model of the simulation vehicle; wherein, the correlation between the historical collision simulation sequence data and the experimental test sequence data is greater than a preset correlation threshold.

[0052] This application provides an optimization device for a car collision simulation model. A model building module 21 constructs an initial model of the simulated vehicle based on its parameters. A preliminary training module 22 inputs historical collision simulation sequence data into the initial model for preliminary training, resulting in a pre-trained model. A retraining module 23 inputs experimental test sequence data into the pre-trained model for retraining, resulting in a target model of the simulated vehicle. The correlation between the historical collision simulation sequence data and the experimental test sequence data is greater than a preset correlation threshold. Specifically, by constructing an initial model of the simulated vehicle and using historical collision simulation sequence data for preliminary training (using a large amount of simulation data for rapid training), the vehicle model quickly reaches a certain level of accuracy. Further training using experimental test sequence data (using a smaller amount of test data) improves the accuracy of the vehicle model. Furthermore, by selecting historical collision simulation sequence data with a high correlation to the experimental test sequence data, the training accuracy of the initial model is improved, thereby enhancing the efficiency and effectiveness of model training and optimization.

[0053] In one embodiment, the aforementioned preliminary training module 22 can be further configured to: input historical collision simulation sequence data into the initial model to obtain a first output result; calculate a first loss value based on the first output result and the observation value corresponding to the historical collision simulation sequence data; if the first loss value is greater than a preset first loss value threshold, adjust the parameters of the initial model until the first loss value is less than or equal to the first loss value threshold to obtain the initial training model.

[0054] In one embodiment, the aforementioned preliminary training module 22 may be further configured to: calculate the objective function values ​​of multiple data in the historical collision simulation sequence data; wherein the objective function represents the first loss value of the corresponding data; calculate the optimal objective function value, the mean of the objective function value, and the standard deviation of the objective function value of the historical collision simulation sequence data; calculate the parameter adjustment step size based on the optimal objective function value, the mean of the objective function value, and the standard deviation of the objective function value; and adjust the parameters of the initial model based on the parameter adjustment step size.

[0055] In one embodiment, the aforementioned preliminary training module 22 can be further configured such that the formula for calculating the parameter adjustment step size is:

[0056] ;

[0057] in, Adjust the step size for the parameters. For regularization functions, The mean of the objective function values. To obtain the optimal objective function value, The standard deviation of the objective function value. The threshold value is set.

[0058] In one embodiment, the historical collision simulation sequence data includes multiple data sets; wherein, the aforementioned preliminary training module 22 can be further configured to: calculate the correlation between each historical collision simulation sequence data set and the experimental test sequence data; if the correlation is greater than the correlation threshold, select the corresponding historical collision simulation sequence data set as the target collision simulation sequence data set; input the target collision simulation sequence data set into the initial model to perform preliminary training on the initial model, thereby obtaining the initial training model.

[0059] In one embodiment, the retraining module 23 can be further configured to: input experimental test sequence data into the initial training model to obtain a second output result; calculate a second loss value based on the second output result and the observation value corresponding to the experimental test sequence data; if the second loss value is greater than a preset second loss value threshold, adjust the parameters of the initial training model until the second loss value is less than or equal to the second loss value threshold to obtain the target model.

[0060] In one embodiment, the retraining module 23 can be further configured to: adjust the parameter values ​​of the initial training model; calculate the parameter error variable of the initial training model; and calculate the parameter error of the target model based on the parameter error and parameter error variable of the initial training model.

[0061] Below, for reference Figure 3 This application describes an electronic device according to embodiments thereof. The electronic device may be either or both of a first device and a second device, or a standalone device independent of them, which may communicate with the first device and the second device to receive acquired input signals from them.

[0062] Figure 3 A block diagram of an electronic device according to an embodiment of this application is illustrated.

[0063] like Figure 3 As shown, the electronic device 10 includes one or more processors 11 and memory 12.

[0064] The processor 11 may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 10 to perform desired functions.

[0065] The memory 12 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 11 may execute the program instructions to implement the methods of the various embodiments of this application described above and / or other desired functions. Various contents such as input signals, signal components, and noise components may also be stored in the computer-readable storage medium.

[0066] In one example, the electronic device 10 may also include an input device 13 and an output device 14, which are interconnected via a bus system and / or other forms of connection mechanism (not shown).

[0067] When the electronic device is a standalone device, the input device 13 can be a communication network connector for receiving the collected input signals from the first device and the second device.

[0068] In addition, the input device 13 may also include, for example, a keyboard, a mouse, etc.

[0069] The output device 14 can output various information to the outside, including determined distance information, direction information, etc. The output device 14 may include, for example, a display, a speaker, a printer, and a communication network and its connected remote output devices, etc.

[0070] Of course, for the sake of simplicity, Figure 3 Only some of the components of the electronic device 10 relevant to this application are shown in this illustration; components such as buses, input / output interfaces, etc., are omitted. In addition, the electronic device 10 may include any other suitable components depending on the specific application.

[0071] In addition to the methods and apparatus described above, embodiments of this application may also be computer program products, which include computer program instructions that, when executed by a processor, cause the processor to perform the steps in the methods according to various embodiments of this application described in the "Exemplary Methods" section above.

[0072] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this application. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0073] Furthermore, embodiments of this application may also be computer-readable storage media storing computer program instructions thereon, which, when executed by a processor, cause the processor to perform the steps in the methods according to various embodiments of this application described in the "Exemplary Methods" section above.

[0074] The computer-readable storage medium may be any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0075] The basic principles of this application have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this application are merely examples and not limitations, and should not be considered as essential features of each embodiment of this application. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the application to the necessity of employing the aforementioned specific details for implementation.

[0076] The block diagrams of devices, apparatuses, devices, and systems involved in this application are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0077] It should also be noted that in the apparatus, equipment, and methods of this application, the components or steps can be disassembled and / or recombined. These disassemblies and / or recombinations should be considered as equivalent solutions of this application.

[0078] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of this application. Therefore, this application is not intended to be limited to the aspects shown herein, but rather to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0079] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this application to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

Claims

1. An optimization method for a car collision simulation model, characterized in that, include: Based on the vehicle parameters of the simulated vehicle, an initial model of the simulated vehicle is constructed; The historical collision simulation sequence data is input into the initial model to perform preliminary training on the initial model, resulting in the initial trained model. The test sequence data is input into the initial training model to retrain the initial training model, thereby obtaining the target model of the simulated vehicle; wherein the correlation between the historical collision simulation sequence data and the test sequence data is greater than a preset correlation threshold. The step of inputting historical collision simulation sequence data into the initial model to perform preliminary training on the initial model and obtaining the initial trained model includes: Input the historical collision simulation sequence data into the initial model to obtain the first output result; Based on the first output result and the observation value corresponding to the historical collision simulation sequence data, the first loss value is calculated; If the first loss value is greater than the preset first loss value threshold, then the parameters of the initial model are adjusted until the first loss value is less than or equal to the first loss value threshold, thus obtaining the initial training model. The adjustment of the parameters of the initial model includes: Calculate the objective function values ​​of multiple data points in the historical collision simulation sequence data; wherein the objective function represents the first loss value of the corresponding data. Calculate the optimal objective function value, the mean of the objective function value, and the standard deviation of the objective function value for the historical collision simulation sequence data; The parameter adjustment step size is calculated based on the optimal objective function value, the mean of the objective function value, and the standard deviation of the objective function value; The parameters of the initial model are adjusted based on the step size of the parameters. The calculation of the parameter adjustment step size based on the optimal objective function value, the mean of the objective function value, and the standard deviation of the objective function value includes: The formula for calculating the parameter adjustment step size is: ； in, Adjust the step size for the parameters. For regularization functions, The mean of the objective function values. To obtain the optimal objective function value, The standard deviation of the objective function value. The threshold value is set.

2. The optimization method for the automobile collision simulation model according to claim 1, characterized in that, The historical collision simulation sequence data includes multiple data sets; wherein, the step of inputting the historical collision simulation sequence data into the initial model to perform preliminary training on the initial model to obtain the initial trained model includes: Calculate the correlation between each of the historical collision simulation sequence data and the experimental test sequence data; If the correlation is greater than the correlation threshold, then the corresponding historical collision simulation sequence data is selected as the target collision simulation sequence data; The target collision simulation sequence data is input into the initial model to perform preliminary training on the initial model, thereby obtaining the initial trained model.

3. The optimization method for the automobile collision simulation model according to claim 1, characterized in that, The step of inputting the test sequence data into the initial training model to retrain the initial training model and obtain the target model of the simulation vehicle includes: The experimental test sequence data is input into the initial training model to obtain the second output result; Based on the second output result and the observed values ​​corresponding to the experimental test sequence data, the second loss value is calculated; If the second loss value is greater than the preset second loss value threshold, then the parameters of the initial training model are adjusted until the second loss value is less than or equal to the second loss value threshold, thus obtaining the target model.

4. The optimization method for the automobile collision simulation model according to claim 3, characterized in that, The adjustment of the parameters of the initial training model includes: Adjust the parameter values ​​of the initial training model; Calculate the parameter error variables of the initial training model; Based on the parameter error of the initial training model and the parameter error variable, the parameter error of the target model is calculated.

5. An optimization device for a car collision simulation model, characterized in that, include: The model building module is used to build an initial model of the simulated vehicle based on the vehicle parameters of the simulated vehicle. The preliminary training module is used to input historical collision simulation sequence data into the initial model to perform preliminary training on the initial model and obtain the initial trained model. The retraining module is used to input the test sequence data into the initial training model to retrain the initial training model and obtain the target model of the simulated vehicle; wherein the correlation between the historical collision simulation sequence data and the test sequence data is greater than a preset correlation threshold. The initial training module is configured as follows: Input the historical collision simulation sequence data into the initial model to obtain the first output result; Based on the first output result and the observation value corresponding to the historical collision simulation sequence data, the first loss value is calculated; If the first loss value is greater than the preset first loss value threshold, then the parameters of the initial model are adjusted until the first loss value is less than or equal to the first loss value threshold, thus obtaining the initial training model. The initial training module is configured as follows: Calculate the objective function values ​​of multiple data points in the historical collision simulation sequence data; wherein the objective function represents the first loss value of the corresponding data. Calculate the optimal objective function value, the mean of the objective function value, and the standard deviation of the objective function value for the historical collision simulation sequence data; The parameter adjustment step size is calculated based on the optimal objective function value, the mean of the objective function value, and the standard deviation of the objective function value; The parameters of the initial model are adjusted based on the step size of the parameters. The initial training module is configured as follows: The formula for calculating the parameter adjustment step size is: ； in, Adjust the step size for the parameters. For regularization functions, The mean of the objective function values. To obtain the optimal objective function value, The standard deviation of the objective function value. The threshold value is set.

6. A computer-readable storage medium, characterized in that, The storage medium stores a computer program for performing the method described in any one of claims 1-4.

7. An electronic device, characterized in that, include: processor; Memory used to store the processor's executable instructions; The processor is used to execute the method described in any one of claims 1-4.

Images