LAUSR

This paper demonstrates a conjoint method integrating the proposed Hybrid Contribution Analysis (HCA) method, the Artificial Neutral Network (ANN) meta-model, the modified Non-dominated Soring Genetic Algorithm II (MNSGAII) and the Ideal Point Method (IPM), used for multi-objective lightweight and crashworthiness optimization of the side structure of an automobile body. First of all, the static-dynamic stiffness models of the automobile body and the vehicle side crashworthiness model are separately established and validated against corresponding actual experiments. Next, the initially selected parts for optimization are screened using the proposed HCA method to determine the final parts for optimization, thicknesses of which are taken as design variables. After that, design of experiment (DoE) coupled with ANN-based meta-models are utilized to approximate the output performance indicators of the automobile body, based on which the modified NSGA-II (MNSGAII) with ε-elimination technique is then employed to solve the multi-objective optimization process, considering the total mass and the torsional stiffness of the automobile body, the maximum intrusion deformation of the measuring point P1 on the inner panel of B-pillar and the measuring point D1 on the inner panel of front door as four optimization objectives. Finally, the IPM method identifies the optimal trade-off solution from the obtained Pareto set, and a comprehensive comparison between the optimized design and the baseline design further confirms the validity of the proposed conjoint method. Specially, the four-objective Pareto set approximately embodies that of each pair of separately run two-objective optimization, thus providing more optimization schemes for designers.