Vehicle collision simulation optimization method based on machine learning

Abstract

The invention discloses a vehicle collision simulation optimization method based on machine learning. The method comprises the following steps: S1, establishing a whole vehicle collision simulation analysis model; s2, establishing a parameterized file of a design variable, and establishing a corresponding data acquisition workflow; s3, obtaining a calculation result file and sampling data; s4, acquiring collision sampling data containing a collision deformation label; s5, determining value ranges of deformation labels, sample points and design variables participating in machine learning model training; s6, performing machine learning model training; s7, judging whether the prediction precision of the trained machine learning model meets a set threshold value or not and judging whether the number of sample points meets a requirement or not, and if both are met, performing a step S8; otherwise, returning to the step S3; s8, establishing a corresponding numerical optimization workflow; s9, performing numerical optimization by adopting an optimization algorithm; and S10, if the numerical value optimization result does not meet the set numerical value optimization precision, increasing the number of the sample points, returning to the step S3, and if the numerical value optimization result meets the set numerical value optimization precision, ending the step.

Description

technical field [0001] The invention relates to a simulation method, in particular to a vehicle collision simulation optimization method. Background technique [0002] In recent years, in order to improve vehicle crash safety performance and save vehicle development costs, domestic and foreign OEMs generally use computer-aided engineering (CAE, Computer Aided Engineering) to perform crash simulation verification and optimization of body structures in the early stage of vehicle development. [0003] In the process of collision safety development of the body structure, if the result of the vehicle collision simulation calculation does not meet the collision safety index, the development engineer needs to analyze the collision simulation calculation result first, and then design optimization measures (such as changing the sheet metal of the body structure) based on the engineering experience. Gold thickness, modify the shape of the body structure and replace the material type o...

AI Technical Summary

Problems solved by technology

[0010] The purpose of the present invention is to provide a vehicle collision simulation optimization method based on machine learning. This method aims at the problems of insufficient accuracy of the current collision simulation numerical optimization method and the incomplete evaluation method of vehicle body structure deformation, and builds a new set of vehicle collision simulation optimization. Process, which can quickly and accurately reduce the optimization space, and finally quickly obtain a reasonable and effective optimization solution

Images