CN114626143A - Automobile collision analysis optimization method, electronic device and storage medium - Google Patents

Abstract

The invention discloses an automobile collision analysis and optimization method, an electronic device and a storage medium, and belongs to the field of automobile collision analysis and optimization. The invention utilizes the parameterized model to automatically drive parameters to update the network characteristics, uses the optimized Latin hypercube algorithm to select the collision sample points (DOE test design), uses the high-performance computing platform and collision simulation software to solve the collision sample points, and then uses The response surface algorithm builds a surrogate model, uses the pointer-2 optimization algorithm to automatically optimize based on the surrogate model, and finally quickly verifies the optimal results through the parametric model to automatically find a reasonable body-in-white structure based on the vehicle crash conditions. The collision automatic optimization method proposed by the invention is effective and requires less manual intervention, and can be applied to the structural optimization design of the early forward development of the body-in-white and the partial optimization of the body under the collision condition.

Description

Automobile crash analysis and optimization method, electronic device and storage medium

technical field

The invention belongs to the field of automobile collision analysis and optimization, and in particular relates to an automobile collision analysis and optimization method, an electronic device and a storage medium.

Background technique

Collision safety simulation analysis is now an indispensable link in automobile research and development, and it is also an important part of CAE analysis in automobile research and development. Through finite element simulation analysis, the actual vehicle collision situation when driving on the road is simulated. According to the simulation analysis results Optimize the body structure to protect the passengers in the car and pedestrians on the road.

Nowadays, different vehicle safety evaluation indicators have put forward requirements and challenges to vehicle safety. However, most of the current body structure design is based on engineers' own experience and reference to competing cars to design the body structure, update the body structure according to the problems exposed by the finite element analysis results, and then analyze until the development target requirements are met. Therefore, how to find a more reasonable and effective way to design the body structure, so as to improve the collision safety performance of the entire body, thereby effectively reducing the cost and risk of body development, is a major demand for current body development.

At present, the main process of optimizing the body structure based on the vehicle crash safety conditions is as follows:

(a) The CAD engineer designs the body-in-white of the model to be developed according to the design experience and the structure of the reference car, and the data storage form of the body structure is the CAD model (generally CATIA model);

(b) The CAE engineer discretizes the CAD model provided by the CAD engineer to obtain the CAE mesh model for finite element analysis, and then conducts a simulation analysis based on the specific working conditions of collision safety to obtain the risk area of the structure, so as to propose a proposal for the structure. Suggestions for Improvement;

(c) CAD engineers make targeted improvements to the body structure according to the results of the simulation analysis, and feed back the CAD model to the CAE engineers after the improvements;

(d) CAE engineers then perform finite element simulation analysis on the new structure until the target requirements are met.

To sum up, the traditional vehicle crash analysis and optimization mainly relies on manual work to achieve the optimization work. In order to achieve the target requirements, a large number of program iterations are needed to try and modify the body structure. The modification of the body structure, the establishment of finite element mesh models, simulation analysis, and optimization schemes The proposal needs to be manually implemented by CAD engineers and CAE engineers, and the process is complicated and requires a lot of manpower and time.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide an automobile collision analysis and optimization method, electronic equipment and storage medium, so as to solve the problem that a large number of program iteration attempts and modification of the body structure are required during the collision analysis and optimization of the whole vehicle.

In order to achieve the above object, the present invention provides an automatic analysis and optimization method for automobile collision based on the parametric model technology. This method proposes an automatic optimization method for vehicle collisions based on parametric model technology, combined with DOE analysis, response surface surrogate model algorithm and pointer-2 optimization algorithm. The program iteratively tries and modifies the problem of the body structure.

The traditional model used for collision simulation cannot record variables because it is not a parametric model, so automatic optimization cannot be performed, and it depends on manual operation. The present invention uses parameterized model technology to record variables, and proposes an optimized combination suitable for vehicle collision optimization, which fills the gap in the field.

An automatic collision optimization method based on parametric model technology, the steps of which include:

S1. Establish a parametric model of the body-in-white, and set the parameters of the parametric model.

The main parameters of the parametric model include the position of the control point, the curvature of the line and the shape of the section, etc. The parametric assembly relationship between the parts is established through the mapping relationship. The characteristics of the parametric model are: it can quickly and automatically generate meshes to realize the integration of CAD and CAE; if a certain parameter is changed, the connection relationship of the surrounding parts will automatically change without manual processing; It can be quickly exchanged between them, and the optimal analysis of the topology can be carried out.

S2. Build a collision analysis model and parameter control.

The transformation parameters of the parametric model are controlled by script, and the mesh model (optimization area) under different parameters is output. The parameterized parameters include thickness, section size, moving distance of the part, etc.

When assembling a collision model, the meshes for the optimized parts are imported as separate parts, and the meshes for other non-optimized areas remain unchanged. The optimized area and the non-optimized area use the corresponding file link. This assembly method can ensure that the parametric model can be automatically assembled into a model that can be used for collision calculation after outputting different meshes (optimized regions).

After the collision model is calculated, a result file is generated, and the displacement, acceleration, etc. concerned in the result file are extracted through the script as the optimization target.

S3. Build a DOE analysis loop. The DOE algorithm adopts the optimized Latin hypercube design, uses the optimized platform-driven parametric model to perform collision calculation according to the DOE sample point parameter values, and automatically combines other files required for collision calculation.

The Latin hypercube design has the following advantages: efficient space-filling ability, requiring fewer sample points for the same level of research; ability to fit second-order or more nonlinear relationships. Optimizing Latin Hypercube Designs improves the uniformity of Latin Hypercube designs, making factor and response fits more accurate and realistic.

S4, using a high-performance computing platform to solve the DOE sample.

S5. Build a response surface surrogate model according to the DOE analysis result data, and check whether the accuracy of the surrogate model is appropriate. If the accuracy is not enough, re-adjust the surrogate model parameters or replace the surrogate model algorithm until the surrogate model accuracy is appropriate.

S6. Use the pointer-2 optimization algorithm to set the optimization objectives and constraints based on the proxy model for optimization. The pointer-2 optimization algorithm is suitable for multi-objective global optimization, and can find the globally optimal results in the proxy model. Different constraints and objectives can be replaced to find multiple sets of optimization results.

S7. For multiple sets of optimization results, use a parametric model to quickly analyze and verify the optimization results and screen the optimal results.

An electronic device comprising one or more processors and a memory; one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs being configured to execute the above-mentioned automobile Collision analysis optimization method.

The present invention also provides a computer-readable storage medium, where program codes are stored in the computer-readable storage medium, wherein the above-mentioned vehicle crash analysis and optimization method is executed when the program codes are executed.

Compared with the prior art, the present invention has the following advantages and beneficial effects:

The present invention utilizes the characteristics of the parameterized model to automatically drive the parameter update sub-network, uses the optimized Latin hypercube algorithm to select the collision sample points (DOE test design), uses the high-performance computing platform and collision simulation analysis software to solve the collision sample points, and then The response surface algorithm is used to build a surrogate model, and the pointer-2 optimization algorithm is used to automatically optimize based on the surrogate model, and finally the parameterized model is used to quickly verify the optimal results to automatically find a reasonable body-in-white structure based on the vehicle crash conditions. The collision automatic optimization method proposed by the invention is effective and requires less manual intervention, and can be applied to the structural optimization design of the early forward development of the body-in-white and the partial optimization of the body under the collision condition.

Description of drawings

Fig. 1 is the flow chart of the collision automatic optimization method based on parametric model technology according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of optimized parameter control and collision model construction according to an embodiment of the present invention.

Detailed ways

In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, but not to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

The invention is effective and efficient, and combines the technical advantages of parameterized model parameter-driven structure automatic change of network division, DOE analysis, response surface proxy model construction, pointer-2 optimization algorithm, etc. The optimal combination of section size, beam position, sheet metal thickness and material for collision analysis conditions reduces a lot of manual effort.

Under the premise that the current research and development cycle is shortened and the rhythm is accelerated, the present invention utilizes the above technical combination to achieve maximum reduction of workload, maximum technical advantage, and the most convenient search for collision body structures.

The collision automatic optimization method based on the parametric model technology according to the embodiment of the present invention, as shown in FIG. 1 , includes the following steps:

S1. Establish a parametric model of the body-in-white, and set the parameters of the parametric model.

The parametric model of the body-in-white is a model generated by reconstructing the parameters of the car model, which can achieve rapid parameterization and modularization of CAD and CAE preprocessing integration. The parameters of the parametric model are the parameters in the car body that will affect the crash performance, such as the section size, position, material thickness, material, etc. of the body beam, and the effective parameters are screened out according to the development constraints of the body and the layout space.

S2. Build a collision analysis model, and write a script to control the parameters of the parameterized model and the simulation calculation results.

As shown in Figure 2, the optimization platform updates the parameters of the parametric model by controlling the parameter file, and then generates the corresponding optimized area grid. The optimized area grid automatically builds a collision analysis model with the non-optimized area grid by connecting the model for calculation. . The results of the collision analysis are automatically converted into optimization objectives and constraints through script files, and then passed to the optimization platform. The optimization platform continuously reads and changes parameter files and result files for optimization, thereby automating the entire process.

S3. Build a DOE analysis loop. The DOE algorithm adopts the optimized Latin hypercube, uses the optimized platform to drive the parametric model to perform the collision calculation according to the DOE sample point parameter values, and automatically combines other files required for the collision calculation.

The DOE optimization algorithm adopts the optimized Latin hypercube algorithm. The algorithm sample points are evenly distributed and the calculation amount is small. Under the premise of satisfying the error analysis and building the surrogate model, it can reduce the calculation amount and save the calculation time.

S4, using a high-performance computing platform to solve the DOE sample.

S5. Use the response surface surrogate model algorithm to construct a surrogate model according to the DOE analysis result data, and check whether the accuracy of the surrogate model is appropriate. If the accuracy is not enough, re-adjust the surrogate model parameters or replace the surrogate model algorithm until the surrogate model accuracy is appropriate.

S6. Use the pointer-2 optimization algorithm to set the optimization objectives and constraints based on the proxy model for optimization. The pointer-2 optimization algorithm is suitable for multi-objective global optimization, and can find the globally optimal results in the proxy model. Different constraints and objectives can be replaced to find multiple sets of optimization results.

The optimization model algorithm adopts the pointer-2 algorithm, and tries various combinations of optimization constraints and objectives to find multiple sets of optimization results.

S7. For multiple sets of optimization results, use a parametric model to quickly analyze and verify the optimization results and screen the optimal results.

The rapid analysis and verification of the optimization results adopts the parametric model technology, using the characteristics of the parametric model to be flexible, fast and fast mesh division, and analyzes the appropriate structural design scheme based on the engineering design experience and implements it in the parametric model. The latest actual finite element analysis and verification effectiveness of the program.

The present invention also provides an electronic device comprising one or more processors and a memory; one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs are configured to Execute the vehicle crash analysis optimization method described above.

The present invention also provides a computer-readable storage medium, where program codes are stored in the computer-readable storage medium, wherein the above-mentioned vehicle crash analysis and optimization method is executed when the program codes are executed.

It should be pointed out that, according to the needs of implementation, the various steps/components described in this application may be split into more steps/components, or two or more steps/components or partial operations of steps/components may be combined into new steps/components to achieve the purpose of the present invention.

Those skilled in the art can easily understand that the above are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present invention should be Included in the protection scope of the present invention.

Claims (4)

1. An automobile collision analysis optimization method is characterized by comprising the following steps:

establishing a white body parameterized model, and establishing parameterized assembly relations among parts by the parameterized model through a mapping relation;

building a collision analysis model, and setting an optimization target and constraint by combining the calculation result of the collision analysis model and the requirement of vehicle model development;

recording optimization parameters based on the parameterized model;

building DOE design, wherein an optimized Latin hypercube is adopted in the DOE algorithm, a parameterized model is driven to output an optimized area grid model under different parameters according to DOE sample points, the optimized area grid model is connected with a non-optimized area grid through a connection model, and a model file required by collision calculation is automatically combined and built;

solving the DOE sample;

constructing a response surface proxy model by using a response surface proxy model algorithm according to DOE analysis result data, checking whether the precision of the proxy model meets the requirement, and if the precision DOEs not meet the requirement, readjusting proxy model parameters or replacing the proxy model algorithm until the precision of the proxy model meets the requirement;

optimizing by setting optimization targets and constraints based on the agent model by using a pointer-2 optimization algorithm, and replacing different optimization targets and constraints to search a plurality of groups of optimization results;

and aiming at the multiple groups of optimization results, rapidly analyzing and verifying the optimization results by using the parameterized model and screening the optimal results.

2. The method for analyzing and optimizing automobile collision according to claim 1, characterized in that the parameters of the body-in-white parameterized model are parameters in the automobile body which have influence on collision performance, and comprise thickness, section size and moving distance of parts.

3. An electronic device comprising one or more processors and memory;

one or more programs stored in the memory and configured to be executed by the one or more processors, the one or more programs configured to perform the vehicle crash analysis optimization method of claim 1 or 2.

4. A computer-readable storage medium, in which a program code is stored, wherein the vehicle collision analysis optimization method according to claim 1 or 2 is performed when the program code is executed.

Images