Robust Multi-objective Optimization Of Vehicle Crash Based On Hybrid Surrogate Model

With the economic development,the automotive industry has developed rapidly and the number of car ownership has increased year by year.Therefor,the traffic congestion and traffic accidents are more and more.The issue of automobiles safety has become a great concern problem to the whole society.At present,automotive safety design mainly adopts deterministic design methods,but due to uncertainties of processing technology,material parameters,fixture positioning,and sheet thickness,it will have a direct impact on the performance of automotive panels,and thus affecting the car crash safety.Therefore,the multi-objective robustness optimization design of automobiles key panels is of great significance for improving the safety of vehicle collision.This paper aims at the safety design of automobile frontal crash and studies the multi-objective robustness optimization design method.Firstly,the vehicle crash finite element model was established to simulate the vehicle crash safety.The simulation results of B-pillar peak acceleration,energy curve and foot pedal intrusion were extracted and compared with the experimental results to verify the validity of the finite element model.Second,sensitivity analysis is performed on the energy absorption of the vehicle's front panels and the optimal design variables are selected.Taking the B-pillar peak acceleration,foot pedal intrusion,and the whole vehicle mass as the optimization goals,a multi-objective optimization model for automobile frontal collision was established to conduct optimization design of the automobile body structure panels.In order to ensure the accuracy of the multi-objective optimization model,the influence of different sampling strategies about the accuracy of the surrogate model was studied.Considering different optimization objectives,the accuracy of different surrogate model is studied.A secondary radial basis function surrogate model is established for the B-pillars peak acceleration and the whole vehicle mass,and a Kriging function surrogate model is established for the foot pedal intrusion.Latin hypercube design method is used to select sample points.Finally,a multi-objective robust optimization dual-response surface model and a multi-objective deterministic optimization model were constructed with the same optimization objectives and design variables.The optimal design calculation of the structural panels was performed with the particle swarm optimization algorithm.The optimization results show that the robustness optimization considers the influence of uncertainties,and the optimization effect is less than the deterministic optimization.However,it has a better optimization effect than the original model,and the product quality characteristics are less fluctuating,and the performance is more stable and practical Engineering significance is great.