Abstract: Aiming at the problem of premature convergence and low efficiency of calculation in reliabilitybased multiobjective optimization, a reliabilitybased quadratic optimization method was proposed. According to the unit normal vector of Pareto frontier solution, the weight coefficients of multiple optimization objectives were calculated. Sequential quadratic programming was used to quadratically optimize the results of multiobjective optimization, and reliability analysis was carried out by moment based method. Combined with experimental design, approximate modeling, quadratic optimization algorithm and moment based method, multiobjective reliability based optimization design was carried out for an aluminium alloy new energy vehicle frame with the wall thickness of parts as design variables, the firstorder mode frequency of bending and torsion and the bending stiffness as constraints, the quality and the torsion stiffness of aluminium alloy frame as optimization objectives. The results indicate that the optimal solution is located on the constraint boundary and the problem of premature multiobjective convergence based on reliability is solved. The moment based method improves the optimization efficiency while guaranteeing the accuracy of reliability optimization. Compared with the original design, the torsional stiffness is increased by 12.96%, and the total mass is decreased by 5.41 kg.
Key words: aluminum alloy frame; moment based method; multiobjective optimization; quadratic optimization method