CN-121980871-A - Column collision sliding table test boundary parameter optimization method

CN121980871ACN 121980871 ACN121980871 ACN 121980871ACN-121980871-A

Abstract

The invention relates to the technical field of automobile safety performance development, in particular to a boundary parameter optimization method for a column collision sliding table test. The method comprises the steps of extracting relevant parts from a finite element collision model for protecting passengers of a whole vehicle, obtaining a simplified model, establishing a simulated sliding table for testing the finite element model, carrying out parameterization setting on key parameters in the model, setting initial values and value ranges of the parameters, establishing a multi-objective optimization function, setting optimized constraint conditions, generating DOE matrixes of a plurality of groups of samples in the value ranges of the parameters, wherein each group of samples comprises all combinations of the parameters, constructing a reduced-order prediction model for predicting rib compression responses under different parameter combinations, and obtaining the optimal parameter combinations meeting the constraint conditions by taking multi-objective function minimization as a target. According to the technical scheme, the calculation complexity can be reduced, the input boundary condition that the side column collides with the sliding table and invades the equipment can be rapidly determined, and the injury condition of passengers under the column collision working condition of the whole car can be highly recovered.

Inventors

JIN YUNFEI
LUO QIAO
Song Saifei
TAI YANG
FAN TIQIANG
LI SHANGSHU
HE ENZE
ZHANG PENG

Assignees

中国汽车工程研究院股份有限公司

Dates

Publication Date: 20260505
Application Date: 20260128

Claims (9)

1. The method for optimizing the boundary parameters of the column collision sliding table test is characterized by comprising the following steps of: extracting relevant parts from a finite element collision model for protecting passengers in the whole vehicle column to obtain a simplified model; Establishing a simulation column collision sliding table test finite element model, wherein the model comprises a sliding table, an upper/middle/lower invasion cylinder and a dummy, and setting boundary conditions and initial conditions corresponding to the model, wherein the boundary conditions and the initial conditions comprise a sliding table loading waveform, invasion behaviors of the invasion cylinder and initial positions and postures of the dummy; performing parameterization setting on key parameters in the model, and setting initial values and value ranges of the parameters; Establishing a multi-objective optimization function, representing the difference between the predicted value of the damage parameters of the upper rib, the middle rib and the lower rib of the dummy chest, which are output by the column collision sliding table test, and the test value of the corresponding parameters in the whole car column collision test data, and setting an optimized constraint condition; Generating DOE matrixes of a plurality of groups of samples in the parameter value range, wherein each group of samples comprises all combinations of key parameters, substituting each group of sample parameters into a column collision sliding table test finite element model for calculation and analysis, and obtaining responses under each group of parameters to form a DOE sample set; constructing a reduced-order prediction model based on the DOE sample set, wherein the reduced-order prediction model is used for predicting rib compression response under different parameter combinations; Performing iterative optimization on the reduced order model by using an optimization algorithm with the minimum of the multiple objective functions as a target to obtain an optimal parameter combination meeting constraint conditions; substituting the optimal parameter combination obtained by predicting the reduced-order prediction model into a column collision sliding table test finite element model for calculation and analysis, and verifying the accuracy of the prediction model.
2. The method for optimizing the boundary parameters of the column collision sliding table test according to claim 1, wherein the method is characterized in that related parts are extracted from a whole vehicle column collision passenger protection finite element collision model to obtain a simplified model, and comprises the following steps: Related parts include WorldSID 50 th type side collision dummy, seat, safety belt, SAB air bag and vehicle door inner decoration assembly, and the detailed grid of the critical parts of the chest and rib of the dummy is reserved, and the geometric cleaning and equivalent mass replacement are carried out on the non-critical parts.
3. The column collision sliding table testing boundary parameter optimization method according to claim 1, wherein key parameters in the model comprise sliding table loading waveform scaling factors, upper/middle/lower invasive cylinder waveform scaling factors and impact cylinder Y-direction offset distances.
4. The column collision sliding table test boundary parameter optimization method according to claim 1, wherein the multi-objective optimization function formula is as follows: In the formula, 、、 Respectively representing the root mean square error of the upper rib of the dummy chest, the root mean square error of the middle rib of the chest and the root mean square error of the lower rib of the chest of the simulation data and the test data; l、、 respectively representing the compression amounts of the upper rib, the middle rib and the lower rib of the chest of the dummy corresponding to each time ti measured in the test, 、、 Respectively representing compression amounts of the upper rib, the middle rib and the lower rib of the chest of the dummy corresponding to each time t i of the simulation calculation, wherein n represents the number of data; The constraints for the optimization are as follows: the optimization target is to enable the root mean square error of the test predicted value of the multi-target optimization sliding table and the test data value of the whole vehicle to be not more than 5.
5. The column collision sliding table testing boundary parameter optimization method according to claim 1, wherein the method for constructing a reduced prediction model by adopting an inherent generalized decomposition model reduced method comprises the following steps: creating a snapshot of each selected time and each result from the simulation results, constructing a snapshot matrix: The basis and modality are calculated by the snapshot matrix and truncated: Constructing a reduced order model with a reduced basis: 。
6. The method for optimizing the boundary parameters of the column collision sliding table test according to claim 1, wherein the predictive model formula is: Wherein, the Is a variable of the physical field and, Is a spatial coordinate of the object to be measured, It is the time that is required for the device to be in contact with the substrate, Is a vector of the parameters which, Representing a spatial pattern, Representing a temporal pattern; representing a parameter pattern; Representing the number of modes.
7. The method for optimizing the boundary parameters of the column collision sliding table test according to claim 1, wherein the iterative optimization of the reduced order model with the objective of minimizing the multiple objective functions is performed by using an optimization algorithm, and the method comprises the following steps: And (5) performing iterative optimization on the model parameters by adopting a gradient descent or genetic algorithm.
8. The method for optimizing the boundary parameters of the column touch sliding table test according to claim 1, further comprising substituting the optimized parameters into a column touch sliding table finite element model for calculation and verification, and comparing whether the root mean square error of the response curve of the model after optimization and the root mean square error of the test curve meet target requirements.
9. The method for optimizing the boundary parameters of the column collision sliding table test according to claim 8, wherein, Substituting the optimal parameter combination into a sliding table finite element model, and outputting compression quantity-time curves of upper, middle and lower ribs of the chest of the dummy; calculating root mean square errors of the three curves and corresponding curves of the whole vehicle column collision test; Verifying whether constraint conditions are satisfied: If not, re-optimizing until all errors meet the requirements.

Description

Column collision sliding table test boundary parameter optimization method Technical Field The invention relates to the technical field of automobile safety performance development, in particular to a boundary parameter optimization method for a column collision sliding table test. Background In the scientific and engineering fields, many problems are difficult to solve due to their numerical complexity or specific constraints. Taking a high-dimensional model as an example, with the increasing of dimensions, the computational complexity rises exponentially and rapidly, bringing huge pressure to computational resources and making the solving process extremely time-consuming and difficult. Meanwhile, many practical problems need to realize multi-parameter collaborative optimization, not only is an accurate optimization matrix required to be established, but also a severe challenge is provided for the calculation efficiency, and how to rapidly and efficiently complete multi-parameter optimization on the premise of ensuring the calculation accuracy becomes a difficult problem to be overcome. In the field of automobile safety research and development, side column collision is taken as a severe side collision working condition in main stream safety regulations (such as Euro-NCAP and C-NCAP) at home and abroad, and is always a key direction of important research and development of automobile host factories. The complexity of the side pillar collision working condition is reflected in a plurality of aspects, and the side pillar collision working condition is closely related to a plurality of factors such as invasion amount, invasion speed, deformation of a vehicle door inner plate and the like, and the factors are mutually interwoven and mutually influenced, so that the injury condition of passengers in the collision process is jointly determined. At present, the column collision sliding table test is used as an emerging test method, and has great potential in the aspect of the optimization of the passenger protection matching under the column collision working condition. The method can effectively reduce the cost, remarkably improve the research and development efficiency, and provides a more economic and efficient way for automobile safety research and development. However, existing related side column intrusion devices have significant drawbacks in terms of key parameters. Specifically, key information such as input parameters and offset distance of an impact cylinder is not clear, so that the simulation effect of the injury condition of a dummy is not satisfactory when the column collision working condition of the vehicle is reproduced by using the column collision sliding table test. Because the damage suffered by the dummy in the actual collision cannot be accurately simulated, the test result is difficult to effectively guide the optimization and improvement of the side column collision trolley test, and the requirements of automobile safety research and development on accurate simulation and effective optimization cannot be met. In the vehicle test, in order to simulate the side collision test of the whole vehicle more truly, the sliding table needs to bear two important tasks, namely, on one hand, the acceleration of the vehicle body is simulated so as to simulate the influence of inertial load of the vehicle body on passengers, and on the other hand, the acceleration of the vehicle door is simulated so as to simulate the damage of the invasion of the vehicle door on the passengers. However, in the actual testing process of the column collision sliding table, the side passenger cabins are generally simplified to install, and the built sliding table testing environment is obviously different from the whole vehicle testing environment. Particularly in terms of the state of motion and the quality of the test piece, it is difficult to maintain a high degree of consistency between the two. The difference enables a large deviation between a sliding table test result and a whole car test result, and how to enable a final test effect and a dummy injury condition of the sliding table test to be in accordance with the height of the whole car becomes a problem faced by the current column collision sliding table test. Disclosure of Invention The invention aims to provide a method for optimizing a column collision sliding table test boundary parameter, which can reduce the calculation complexity, quickly determine the input boundary condition of a side column collision sliding table intrusion device and highly reappear the injury condition of passengers under a column collision working condition of a vehicle. In order to achieve the above purpose, the present invention provides a method for optimizing a boundary parameter of a column collision sliding table test, including: extracting relevant parts from a finite element collision model for protecting passengers in the whole vehicle column to obtain a simplified model;