The Reliability-based Design Optimization For Lightweight Design Of Vehicle Body Based On Sequential Kriging Model

With the rapid development of the automobile industry,the problem of energy shortage and environmental pressure has become increasingly prominent.The weight of vehicle body accounted for about 40 percent of vehicle weight,therefore,it is important for energy conservation and environmental protection to reduce the weight of automobile.In engineering application,due to the manufacturing process and other aspects of the uncertainties,the uncertainty of these random variables directly affects the structural performance of the vehicle,,resulting in the lack of reliability of the lightweight design and losing the practical engineering feasibility.Therefore,it is very meaningful for the reliability-based design optimization for lightweight of vehicle body..However,the vehicle collision is a highly nonlinear process.and the direct call of the finite element model for reliability optimization design will spend a lot of computational costs.The use of approximate model technology can effectively reduce the amount of calculation and the computational cost.However,it is difficult to obtain satisfactory results by using the one time modeling and optimization.By using the method of sequence optimization,the optimization method can improve the accuracy of the approximate model,so as to enhancing the effect of optimization design.In this paper,a reliability-based design optimization based on sequential Kriging model is proposed.The Kriging approximation model of the objective and the constraint functions are constructed by Latin square sampling.a performance measure analysis is adopted to evaluate the probability constraints,The nested optimization is transformed into single level optimization problem by using the sequential optimization and reliability assessment.At the each iteration of optimization,new sample points are added around the neighborhood of approximate by using relative improvement criterion,the surrogate model is constantly updated to enhance the fitting accuracy in the region of interest and the global prediction ability of the model.This treatment significantly improve the accuracy of reliability analysis at the optimum design point.The numerical example shows that the accuracy of the approximate model is greatly improved.The method is then applied to the lightweight design of vehicle body structure,The lightweight of vehicle body was considered as objective,the reliability indices for safety performance of vehicle frontal 100%impact is chosen as constraints.and the thickness of the front end main energy absorbing component is selected as the design variable.The results show that the lightweight of the selected parts of the vehicle body can meet the requirements of the safety and reliability of the vehicle,while weight of the vehicle body is reduced the 11.39kg.