Research On Parameterization Optimization Design Of Vehicle Structure For Automotive Pedestrian Protection

Because of the huge number of automobiles,bicycles and motorcycles in China,the traffic situation is complex and traffic accidents are frequent.Statistics show that the casualties of pedestrians and other vulnerable groups in traffic accidents are more serious than that of drivers,and the safety of pedestrians is greatly threatened.The optimization design of vehicle body structure for pedestrian protection is an important research direction.Modern process of structure optimization design is based on Computer-Aided Engineering and data mining technology.Model substitution test is an important technical means to improve the efficiency of product design.However,the traditional optimization design process can’t meet the needs of complex system problems in terms of model accuracy and information utilization.The pedestrian protection oriented optimization design of body structure is a multi-dimensional,highly non-linear complex system problem.In order to solve the optimization design problem of vehicle structure for pedestrian safety performance,the parameterization optimization design integration process combined with digital technology is built for complex engineering problems.The following aspects are studied in this paper:Firstly,based on the finite element model of the existing vehicle design scheme,the collision model of pedestrian protection is constructed.According to the sensitivity analysis,the components that have significant impact on the performance of vehicle pedestrian protection are found.The shapes and positions of the components’ structure are as parameters.Parameters drive the transformation of the structure with the help of the grid deformation technology,and then the efficient and automatic transformation of the CAE model with different vehicle design schemes is realized.So as to lay the model foundation for further optimization design.Secondly,the surrogate updating strategy based on Bayesian theory is proposed.Through the bias analysis between the responses of initial surrogate model and the simulation results of high fidelity CAE model,the bias model based on Bayesian theory is constructed.On the basis of the initial surrogate model and bias model,a high fidelity surrogate model is modeled to correct the prediction ability of the initial surrogate model.This method can enhance the precision of surrogate model and improve the reliability of the design solution.Then,according to combination of optimization process and data mining technology,a design domain recognition method based on C4.5 algorithm is proposed.Based the efficient design domain reduction algorithm,the feasible domain classification tree based on sample data is obtained,and the feasible region is selected.On the basis of feasible domain,the optimal design improves the convergence efficiency of the optimization objective and the probability of obtaining the optimal solution,so as to reduce the calculation cost and improve the design efficiency.Finally,the process of parameterization optimization design for pedestrian safety performance is formed,which integrates parameterization modeling technology,highprecision surrogate model correction strategy and design domain identification reduction technology.A vehicle body structure optimization based on pedestrian head safety performance is utilized to demonstrate the validity of the proposed process.The hood is as the main research component,and pedestrian head injury index is as the optimization object to enhance the pedestrian protection performance of the vehicle while keeping the stiffness of the hood.The average value of pedestrian head injury index of the optimized hood is reduced effectively while stiffness performance of that meets constraints.The research results show that the proposed optimization process can meet the needs of pedestrian safety oriented vehicle body structure optimization,which is a complex engineering problem.