Multi-objective optimization method for thickness of anti-collision beam of automobile composite bumper

Abstract

The invention discloses a multi-objective optimization method for the thickness of an anti-collision beam of an automobile composite bumper, and the method comprises the following steps: 1, respectively determining the value ranges of the layer thickness of an upper plate, a lower plate, a front plate, a rear plate and a rib plate of the anti-collision beam, and the layer sequence corresponding to each layer thickness; 2, generating an initial sample data set, and determining an objective function f (x) and a constraint function g (x) of anti-collision beam paving layer thickness optimization; 3, according to the objective function and the constraint function, setting an initial parameter of a Gaussian random process kernel function, and training to obtain a GPR model; 4, obtaining a Pareto front solution set through a GPR model, screening out a new sample from the Pareto front solution set, and adding the new sample into the initial sample data set; and step 4, repeating the step 2 to the step 4 until the number of iterations is met, obtaining an optimized Pareto front solution set, and obtaining the optimized layer thickness of the upper plate, the lower plate, the front plate, the rear plate and the rib plate of the anti-collision beam according to the optimized Pareto front solution set.

Description

technical field [0001] The invention belongs to the technical field of automobile passive safety research, in particular to a multi-objective optimization method for the thickness of an automobile composite material bumper anti-collision beam. Background technique [0002] At present, multi-objective optimization is mostly based on the optimization method of the surrogate model. Before optimization, a surrogate model with sufficient accuracy must be established. Usually, the accuracy of the surrogate model can only be accurate and credible when the accuracy of the surrogate model reaches above 0.9. However, for high-dimensional and highly nonlinear mathematical models, such as car crashes, it is necessary to use hundreds or even thousands of groups of Design of Experiments (DOE) for simulation and simulation. The price is unbearable. In addition, most traditional proxy models do not directly reflect the uncertainty of the model, resulting in large deviations in the optimiza...

AI Technical Summary

Problems solved by technology

the price is unbearable

In addition, most traditional proxy models do not directly reflect the uncertainty of the model, resulting in large deviations in the optimization results. When faced with higher-dimensional optimization, the optimization cannot be carried out smoothly.

In addition, for one-step optimization design, the modeling is one-time, and there is no process of approximate model improvement. If the model has a large approximation error, the optimization based on the model must be inaccurate

At present, there is no optimization method that combines sequential optimization and Gaussian process regression model to achieve multi-objective optimization

Images