Vehicle restraint system optimization method and device

Abstract

The invention discloses a vehicle restraint system optimization method and device, and relates to the technical field of automobiles. According to the method, the constraint system performance corresponding to a plurality of parameter value combinations of the constraint system is obtained through simulation, and then machine learning is carried out on the known constraint system performance and value combinations to obtain an approximate model between the constraint system performance and a plurality of parameters; determining a constraint system parameter value required for realizing a constraint system performance target according to the approximation model, designing the constraint system based on the parameter value to enable the performance of the constraint system to reach the target, and realizing optimization of the constraint system; and after each parameter value of the constraint system is determined, each parameter value is input into the approximation model to predict the performance of the constraint system. The method is mainly used for performance optimization and performance prediction of the vehicle restraint system.

Description

Technical Field

[0001] The present invention relates to the field of automobile technology, and in particular to a vehicle restraint system optimization method and device. Background Art

[0002] The vehicle restraint system mainly includes airbags, seat belts, seats and other components, which are used to protect passengers when the vehicle collides. The various parameters of the restraint system determine the performance of the restraint system. If the parameters are not designed properly, the restraint system will have poor protection effect on the occupants. In this case, the restraint system needs to be optimized. Summary of the invention

[0003] The present invention solves the technical problem of how to optimize a vehicle restraint system by providing a vehicle restraint system optimization method and device.

[0004] On the one hand, the present invention provides the following technical solutions:

[0005] A vehicle restraint system optimization method, comprising:

[0006] Establishing a finite element simulation model of a restraint system, performing DOE test design on multiple parameters of the restraint system to obtain multiple value combinations of the multiple parameters, assigning values ​​to the multiple parameters of the finite element simulation model based on the value combinations, solving the finite element simulation model by finite element solving software, and obtaining performance curves and performance indicators corresponding to each value combination;

[0007] Performing machine learning on multiple groups of value combinations and the performance curves, and multiple groups of value combinations and the performance indicators, to obtain approximate models between the performance curves and the multiple parameters, and between the performance indicators and the multiple parameters;

[0008] The performance curve or the performance index to be optimized is optimized, and a value combination of the multiple parameters corresponding to the optimized performance curve or the performance index is determined according to the approximate model.

[0009] Optionally, the performing machine learning on the multiple groups of value combinations and the performance curve, and the multiple groups of value combinations and the performance indicators to obtain approximate models between the performance curve and the multiple parameters, and between the performance indicator and the multiple parameters, includes:

[0010] Use multiple algorithms to train multiple groups of value combinations and the performance curves, and multiple groups of value combinations and the performance indicators, respectively, to obtain multiple approximate models corresponding to the performance curves and multiple approximate models corresponding to the performance indicators;

[0011] Calculating the accuracy of each of the approximate models corresponding to the performance curve and each of the approximate models corresponding to the performance index;

[0012] Optimizing the algorithm with the highest accuracy corresponding to the performance curve and the algorithm with the highest accuracy corresponding to the performance indicator;

[0013] Use the optimized algorithm with the highest accuracy corresponding to the performance curve to train multiple groups of value combinations and the performance curve to obtain the final approximate model between the performance curve and the multiple parameters;

[0014] The optimized algorithm with the highest accuracy corresponding to the performance indicator is used to train multiple groups of value combinations and the performance indicator to obtain the final approximate model between the performance indicator and the multiple parameters.

[0015] Optionally, the using of multiple algorithms to respectively train multiple groups of the value combinations and the performance curves, and multiple groups of the value combinations and the performance indicators, includes:

[0016] An orthogonal decomposition algorithm, a radial basis algorithm and a Kriging algorithm are used respectively to train the plurality of value combinations and the performance curves.

[0017] Optionally, the using of multiple algorithms to respectively train multiple groups of the value combinations and the performance curves, and multiple groups of the value combinations and the performance indicators, includes:

[0018] A weighted interpolation algorithm, a radial basis algorithm and a Kriging algorithm are used respectively to train the plurality of value combinations and the performance indicators.

[0019] Optionally, the performance curve includes an acceleration curve, and the performance index includes a HIC damage value;

[0020] The optimizing the performance curve or the performance index to be optimized includes:

[0021] The acceleration curve is globally reduced to the original set ratio;

[0022] The HIC damage value is determined according to the updated acceleration curve.

[0023] Optionally, determining the HIC damage value according to the updated acceleration curve includes:

[0024] According to the formula determining the HIC damage value;

[0025] HIC is the HIC damage value, t1 is the recording start time, t2 is the recording end time, and a is the acceleration curve.

[0026] Optionally, after performing machine learning on the multiple groups of value combinations and the performance curves, and the multiple groups of value combinations and the performance indicators to obtain approximate models between the performance curves and the multiple parameters, and between the performance indicators and the multiple parameters, the method further includes:

[0027] Obtaining actual values ​​of the plurality of parameters of the restraint system;

[0028] The actual values ​​of the multiple parameters are input into the approximate model to predict the performance curve and the performance index of the restraint system.

[0029] On the other hand, the present invention also provides the following technical solutions:

[0030] A vehicle restraint system optimization device, comprising:

[0031] A simulation module, used to establish a finite element simulation model of a restraint system, perform DOE test design on multiple parameters of the restraint system to obtain multiple value combinations of the multiple parameters, assign values ​​to the multiple parameters of the finite element simulation model based on the value combinations, solve the finite element simulation model by finite element solving software, and obtain performance curves and performance indicators corresponding to each value combination;

[0032] A modeling module, used for performing machine learning on multiple groups of value combinations and the performance curves, and multiple groups of value combinations and the performance indicators, to obtain approximate models between the performance curves and the multiple parameters, and between the performance indicators and the multiple parameters;

[0033] The optimization module is used to optimize the performance curve or the performance index to be optimized, and determine the value combination of the multiple parameters corresponding to the optimized performance curve or the performance index according to the approximate model.

[0034] On the other hand, the present invention also provides the following technical solutions:

[0035] A computer device comprises a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of any vehicle restraint system optimization method.

[0036] On the other hand, the present invention also provides the following technical solutions:

[0037] A computer-readable storage medium stores a computer program, which implements the steps of any vehicle restraint system optimization method when executed by a processor.

[0038] One or more technical solutions provided by the present invention have at least the following technical effects or advantages:

[0039] The present invention obtains the constraint system performance corresponding to the combination of multiple parameter values ​​of the constraint system through simulation, and then performs machine learning on the known constraint system performance and value combinations to obtain an approximate model between the constraint system performance and multiple parameters. The constraint system parameter values ​​required to achieve the constraint system performance target are determined according to the approximate model. After the constraint system is designed based on the parameter values, the performance of the constraint system can reach the target and the constraint system can be optimized. After the various parameter values ​​of the constraint system are determined, the various parameter values ​​can be input into the approximate model to predict the performance of the constraint system. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0041] Figure 1 A flow chart of a vehicle restraint system optimization method according to an embodiment of the present invention;

[0042] Figure 2 A schematic diagram of an acceleration curve of a verification set in an embodiment of the present invention;

[0043] Figure 3 Schematic diagram of the acceleration curve of the test set in an embodiment of the present invention;

[0044] Figure 4 Schematic diagram of a vehicle restraint system optimization device in an embodiment of the present invention. DETAILED DESCRIPTION

[0045] The embodiments of the present invention solve the technical problem of how to optimize the vehicle restraint system by providing a vehicle restraint system optimization method and device.

[0046] In order to better understand the technical solution of the present invention, the technical solution of the present invention will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0047] like Figure 1 As shown, the vehicle restraint system optimization method of the embodiment of the present invention includes:

[0048] Step S1, establishing a finite element simulation model of the constraint system, performing DOE test design on multiple parameters of the constraint system, obtaining multiple value combinations of the multiple parameters, assigning values ​​to the multiple parameters of the finite element simulation model based on the value combinations, solving the finite element simulation model by finite element solving software, and obtaining performance curves and performance indicators corresponding to each value combination;

[0049] Among them, the restraint system includes airbags, seat belts, seats and other structures that directly contact with the occupants and play a protective role during vehicle collision; the finite element simulation model of the restraint system mainly includes models of seat belts, airbags, dashboards, air conditioners, seats, body structures, carpets, steering systems, door interiors, seats, dummies, etc.

[0050] Among them, multiple parameters are parameters that can be parametrically designed, including seat belt parameters (such as force limit level, ignition time, high adjuster position, etc.), airbag parameters (exhaust hole diameter, drawstring length, ignition time, etc.), seat position parameters, front cabin structure material parameters, material thickness parameters, etc., each parameter has an upper limit and a lower limit;

[0051] Among them, DOE (DESIGN OF EXPERIMENT) experimental design can combine values ​​for multiple parameters. Each group of value combinations is independent of each other and relatively evenly distributed in space. For example, the five groups of value combinations in Table 1 can be used to perform DOE experimental design through the super Latin square method;

[0052] Table 1

[0053]

[0054] Among them, each value combination should be used to assign values ​​to multiple parameters of the finite element simulation model respectively, specifically, the parameter values ​​in the INCLUDE files of each subsystem in the finite element simulation model are changed;

[0055] The finite element solution software may be LS-DYNA, and each time a set of value combinations is assigned, all performance curves and performance indicators of the finite element simulation model are solved once. For example, if there are 5 value combinations, each performance curve and each performance indicator has 5;

[0056] Among them, the performance curves may include acceleration curves, chest compression curves, head synthetic acceleration curves, neck shear force curves, neck tension curves, neck extension bending moment curves, chest compression deformation curves, chest viscosity index curves, knee slip curves, etc., and the performance indicators may include HIC injury value, 3ms synthetic acceleration, neck extension bending moment, neck axial force, neck shear force, chest compression deformation, chest viscosity index, left thigh compression force, knee slip, calf compression force, calf compression force, tibia index, tibia index, etc.;

[0057] Step S2, performing machine learning on multiple sets of value combinations and performance curves, and multiple sets of value combinations and performance indicators, to obtain approximate models between the performance curves and multiple parameters, and between the performance indicators and multiple parameters;

[0058] It can be understood that the approximate model characterizes the relationship between each performance curve and multiple parameters, and the relationship between each performance index and multiple parameters, such as the relationship between the acceleration curve and the four parameters in Table 1, and the relationship between the HIC damage value and the four parameters in Table 1. That is, the approximate model corresponding to each performance curve and each performance index can be obtained. After the performance target of the constraint system is clarified, the performance target can be input into the corresponding approximate model to inversely obtain the parameter values ​​of the constraint system required to achieve the performance target.

[0059] Step S3, optimizing the performance curve or performance index to be optimized, and determining a value combination of multiple parameters corresponding to the optimized performance curve or performance index according to the approximate model;

[0060] For example, after optimizing a certain performance curve, the optimized performance curve can be input into the approximate model corresponding to the performance curve to inversely obtain the value combination of multiple parameters of the constraint system. After the constraint system is designed based on the value combination, the performance curve of the constraint system can be made close to or reach the optimization requirements.

[0061] As can be seen from the above, the vehicle restraint system optimization method of the embodiment of the present invention obtains the restraint system performance curve and performance index corresponding to the combination of multiple parameter values ​​of the restraint system through simulation, and then performs machine learning on the known performance curve and value combination, and the performance index and value combination to obtain the relationship between the restraint system performance curve and multiple parameters, and between the restraint system performance index and multiple parameters. According to the relationship, the restraint system parameter values ​​required to achieve the performance target of the restraint system are determined. After designing the restraint system based on the parameter values, the performance of the restraint system can reach the target, thereby realizing the optimization of the restraint system.

[0062] It can be understood that after obtaining the approximate model corresponding to each performance curve and each performance index, after the values ​​of each parameter of the constraint system are determined, the values ​​of each parameter can be input into the approximate model to predict the performance of the constraint system, that is, after step S2, the vehicle restraint system optimization method of the embodiment of the present invention can also include: obtaining actual values ​​of multiple parameters of the constraint system; inputting the actual values ​​of the multiple parameters into the approximate model to predict the performance curve and performance index of the constraint system.

[0063] In step S2, only one algorithm can be used to train multiple sets of value combinations and performance curves, and multiple sets of value combinations and performance indicators. However, if the accuracy of the algorithm is low, the accuracy of the trained approximate model will be low, resulting in inaccurate predicted constraint system performance or inaccurate values ​​of multiple parameters obtained by inversely calculating the constraint system performance.

[0064] In order to improve the accuracy of the approximate model, step S2 may include: using multiple algorithms to train multiple groups of value combinations and performance curves, multiple groups of value combinations and performance indicators, respectively, to obtain multiple approximate models corresponding to the performance curves and multiple approximate models corresponding to the performance indicators; calculating the accuracy of each approximate model corresponding to the performance curve and each approximate model corresponding to the performance indicator; optimizing the algorithm with the highest accuracy corresponding to the performance curve and the algorithm with the highest accuracy corresponding to the performance indicator; using the optimized algorithm with the highest accuracy corresponding to the performance curve to train multiple groups of value combinations and performance curves, to obtain the final approximate model between the performance curve and multiple parameters; using the optimized algorithm with the highest accuracy corresponding to the performance indicator to train multiple groups of value combinations and performance indicators, to obtain the final approximate model between the performance indicator and multiple parameters. This is equivalent to using multiple algorithms to train the approximate model, optimizing the algorithm with the highest accuracy, and then retraining with the optimized algorithm with the highest accuracy, so that the obtained approximate model has higher accuracy, which is conducive to improving the prediction accuracy of the constraint system performance and the accuracy of the constraint system parameter values ​​obtained.

[0065] In step S2, multiple algorithms are used to train multiple groups of value combinations and performance curves, and multiple groups of value combinations and performance indicators, respectively, which may include: using an orthogonal decomposition algorithm, a radial basis algorithm, and a kriging algorithm to train multiple groups of value combinations and performance curves. The optimization of the most accurate algorithm among the orthogonal decomposition algorithm, the radial basis algorithm, and the kriging algorithm may be to optimize its interpolation function, basis function, number of models, etc.

[0066] The embodiment of the present invention divides the value combination of multiple parameters, performance curves, and performance index data into a training set, a validation set, and a prediction set. Step S2 uses the training set data for training. Taking the performance curve as an acceleration curve as an example, the acceleration curve of the validation set is verified using the approximate model trained by the optimized algorithm with the highest accuracy. The results are as follows: Figure 2 As shown, it can be seen that the acceleration curve of the validation set is closer to the acceleration curve obtained by simulation; the approximate model trained with the most accurate optimized algorithm is used to predict the acceleration curve of the prediction set. The results are as follows Figure 3 As shown, it can be seen that the acceleration curve of the predicted set is closer to the acceleration curve obtained by simulation. This proves that the approximate model trained by the optimized algorithm has higher accuracy.

[0067] In step S2, multiple algorithms are used to train multiple groups of value combinations and performance curves, multiple groups of value combinations and performance indicators, respectively, which may include: using weighted interpolation algorithm, radial basis algorithm and Kriging algorithm to train multiple groups of value combinations and performance indicators respectively. The optimization of the algorithm with the highest accuracy among the weighted interpolation algorithm, radial basis algorithm and Kriging algorithm can also be to optimize its interpolation function, basis function, number of models, etc. For example, the highest accuracy algorithm and accuracy corresponding to multiple performance indicators are shown in Table 2, and the accuracy is more than 95%.

[0068] Table 2

[0069] Performance Indicators Highest accuracy algorithm <![CDATA[Precision R 2 > HIC damage value Weighted interpolation 97.69％ 3ms synthetic acceleration Krieger 99.32％ Neck extension moment Radial Basis 95.08％ Neck axial force Radial Basis 99.38％

[0070] In the embodiment of the present invention, after optimizing the constraint system parameter values, not only the optimized performance of the constraint system will change, but also the unoptimized performance will change. The requirement for the constraint system is that all performance must be within the required range and not lower than a certain performance lower limit. Therefore, after optimizing a certain performance, it is necessary to determine whether other performances meet the requirements, so that when other performances do not meet the requirements, the constraint system is optimized again until all performances meet the requirements. To this end, after step S3, the constraint system optimization method also includes: inputting the optimized value combination into the approximate model corresponding to each unoptimized performance curve and each unoptimized performance index, and obtaining each performance curve and each performance index corresponding to the optimized value combination; and determining whether each performance curve and each performance index corresponding to the optimized value combination are within the required range.

[0071] In the embodiment of the present invention, the performance curve is an acceleration curve and the performance index is a HIC damage value. The optimization of the performance curve or performance index to be optimized may include: globally reducing the acceleration curve to the original set ratio; determining the HIC damage value according to the updated acceleration curve. The set ratio may be 90%, and determining the HIC damage value according to the updated acceleration curve includes: according to the formula Determine the HIC damage value; HIC is the HIC damage value, t1 is the start time of recording, t2 is the end time of recording, t2-t1≤15ms, a is the acceleration curve. The results show that after optimizing the acceleration curve, the HIC damage value can be reduced to 80% of the original value.

[0072] like Figure 4 As shown, the embodiment of the present invention also provides a vehicle restraint system optimization device, including:

[0073] The simulation module is used to establish a finite element simulation model of the constraint system, perform DOE test design on multiple parameters of the constraint system, obtain multiple value combinations of the multiple parameters, assign values ​​to multiple parameters of the finite element simulation model based on the value combinations, solve the finite element simulation model through finite element solving software, and obtain performance curves and performance indicators corresponding to each value combination;

[0074] A modeling module is used to perform machine learning on multiple sets of value combinations and performance curves, and multiple sets of value combinations and performance indicators, to obtain approximate models between the performance curve and multiple parameters, and between the performance indicator and multiple parameters;

[0075] The optimization module is used to optimize the performance curve or performance index to be optimized, and determine the value combination of multiple parameters corresponding to the optimized performance curve or performance index according to the approximate model.

[0076] Furthermore, the modeling module can be specifically used to: use multiple algorithms to train multiple groups of value combinations and performance curves, and multiple groups of value combinations and performance indicators, respectively, to obtain multiple approximate models corresponding to the performance curves and multiple approximate models corresponding to the performance indicators; calculate the accuracy of each approximate model corresponding to the performance curve and each approximate model corresponding to the performance indicator; optimize the algorithm with the highest accuracy corresponding to the performance curve and the algorithm with the highest accuracy corresponding to the performance indicator; use the optimized algorithm with the highest accuracy corresponding to the performance curve to train multiple groups of value combinations and performance curves to obtain the final performance curve and the approximate model between multiple parameters; use the optimized algorithm with the highest accuracy corresponding to the performance indicator to train multiple groups of value combinations and performance indicators to obtain the final approximate model between the performance indicator and multiple parameters.

[0077] Furthermore, the modeling module can be specifically used to train multiple groups of value combinations and performance curves using an orthogonal decomposition algorithm, a radial basis algorithm, and a kriging algorithm, respectively.

[0078] Furthermore, the modeling module can be specifically used to train multiple groups of value combinations and performance indicators using weighted interpolation algorithm, radial basis algorithm and Kriging algorithm respectively.

[0079] Further, the performance curve includes an acceleration curve, and the performance index includes a HIC damage value;

[0080] The optimization module can be specifically used to: globally reduce the acceleration curve to the original set ratio; and determine the HIC damage value according to the updated acceleration curve.

[0081] Furthermore, the optimization module can be used to: Determine the HIC damage value; HIC is the HIC damage value, t1 is the recording start time, t2 is the recording end time, and a is the acceleration curve.

[0082] Furthermore, the vehicle restraint system optimization device may also include:

[0083] The prediction module is used to obtain the actual values ​​of multiple parameters of the constraint system; the actual values ​​of the multiple parameters are input into the approximate model to predict the performance curve and performance index of the constraint system.

[0084] Based on the same inventive concept as the vehicle restraint system optimization method described above, an embodiment of the present invention further provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of any of the vehicle restraint system optimization methods described above.

[0085] Since the computer device introduced in the embodiment of the present invention is a computer device used to implement the vehicle restraint system optimization method in the embodiment of the present invention, based on the vehicle restraint system optimization method introduced in the embodiment of the present invention, a person skilled in the art can understand the specific implementation of the computer device in the embodiment of the present invention and its various variations, so how the computer device implements the method in the embodiment of the present invention is not described in detail here. As long as a person skilled in the art implements the computer device used in the vehicle restraint system optimization method in the embodiment of the present invention, it belongs to the scope of protection of the present invention.

[0086] Based on the same inventive concept as the above vehicle restraint system optimization method, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of any of the above vehicle restraint system optimization methods when executed by a processor.

[0087] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0088] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 These computer program instructions can also be stored in a computer-readable memory that can guide a computer or other programmable data processing device to work in a specific way, so that the instructions stored in the computer-readable memory produce a product including an instruction device, which implements the functions specified in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0089] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0090] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0091] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A vehicle restraint system optimization method, characterized in that: include: Establishing a finite element simulation model of a restraint system, performing DOE test design on multiple parameters of the restraint system to obtain multiple value combinations of the multiple parameters, assigning values ​​to the multiple parameters of the finite element simulation model based on the value combinations, solving the finite element simulation model by finite element solving software, and obtaining performance curves and performance indicators corresponding to each value combination; Performing machine learning on multiple groups of value combinations and the performance curves, and multiple groups of value combinations and the performance indicators, to obtain approximate models between the performance curves and the multiple parameters, and between the performance indicators and the multiple parameters; The performance curve or the performance index to be optimized is optimized, and a value combination of the multiple parameters corresponding to the optimized performance curve or the performance index is determined according to the approximate model.

2. The vehicle restraint system optimization method according to claim 1, characterized in that: The performing machine learning on the multiple groups of value combinations and the performance curves, and the multiple groups of value combinations and the performance indicators to obtain approximate models between the performance curves and the multiple parameters, and between the performance indicators and the multiple parameters, includes: Use multiple algorithms to train multiple groups of value combinations and the performance curves, and multiple groups of value combinations and the performance indicators, respectively, to obtain multiple approximate models corresponding to the performance curves and multiple approximate models corresponding to the performance indicators; Calculating the accuracy of each of the approximate models corresponding to the performance curve and each of the approximate models corresponding to the performance index; Optimizing the algorithm with the highest accuracy corresponding to the performance curve and the algorithm with the highest accuracy corresponding to the performance indicator; Use the optimized algorithm with the highest accuracy corresponding to the performance curve to train multiple groups of value combinations and the performance curve to obtain the final approximate model between the performance curve and the multiple parameters; The optimized algorithm with the highest accuracy corresponding to the performance indicator is used to train multiple groups of value combinations and the performance indicator to obtain the final approximate model between the performance indicator and the multiple parameters.

3. The vehicle restraint system optimization method according to claim 2, characterized in that: The using of multiple algorithms to respectively train multiple groups of the value combinations and the performance curves, and multiple groups of the value combinations and the performance indicators, includes: An orthogonal decomposition algorithm, a radial basis algorithm and a Kriging algorithm are used respectively to train the plurality of value combinations and the performance curves.

4. The vehicle restraint system optimization method according to claim 2, characterized in that: The using of multiple algorithms to respectively train multiple groups of the value combinations and the performance curves, and multiple groups of the value combinations and the performance indicators, includes: A weighted interpolation algorithm, a radial basis algorithm and a Kriging algorithm are used respectively to train the plurality of value combinations and the performance indicators.

5. The vehicle restraint system optimization method according to claim 1, characterized in that: The performance curve includes an acceleration curve, and the performance index includes a HIC damage value; The optimizing the performance curve or the performance index to be optimized includes: The acceleration curve is globally reduced to the original set ratio; The HIC damage value is determined according to the updated acceleration curve.

6. The vehicle restraint system optimization method according to claim 5, characterized in that: The determining the HIC damage value according to the updated acceleration curve includes: According to the formula determining the HIC damage value; HIC is the HIC damage value, t1 is the recording start time, t2 is the recording end time, and a is the acceleration curve.

7. The vehicle restraint system optimization method according to claim 1, characterized in that: After performing machine learning on the multiple groups of value combinations and the performance curves, and the multiple groups of value combinations and the performance indicators to obtain approximate models between the performance curves and the multiple parameters, and between the performance indicators and the multiple parameters, the method further includes: Obtaining actual values ​​of the plurality of parameters of the restraint system; The actual values ​​of the multiple parameters are input into the approximate model to predict the performance curve and the performance index of the restraint system.

8. A vehicle restraint system optimization device, characterized in that: include: A simulation module, used to establish a finite element simulation model of a restraint system, perform DOE test design on multiple parameters of the restraint system to obtain multiple value combinations of the multiple parameters, assign values ​​to the multiple parameters of the finite element simulation model based on the value combinations, solve the finite element simulation model by finite element solving software, and obtain performance curves and performance indicators corresponding to each value combination; A modeling module, used for performing machine learning on multiple groups of value combinations and the performance curves, and multiple groups of value combinations and the performance indicators, to obtain approximate models between the performance curves and the multiple parameters, and between the performance indicators and the multiple parameters; The optimization module is used to optimize the performance curve or the performance index to be optimized, and determine the value combination of the multiple parameters corresponding to the optimized performance curve or the performance index according to the approximate model.

9. A computer device, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images