Abstract: With consideration of the influence of uncertain parameters of occupant restraint system on occupant safety, a multiobjective uncertainty optimization technique for occupant restraint system is proposed based on the combination of nonlinear interval number programming (NINP) and approximate model with locallyrefined mesh. The numerical model for occupant restraint system is corrected according to the results of real vehicle frontal crash tests, and the uncertainty optimization problem is converted into a deterministic optimization problem by using interval order relations. Intergeneration projection genetic algorithm (IPGA) and micro multiobjective genetic algorithm (μMOGA) are adopted to seek for Pareto solution set meeting the protection performance requirements of occupant restraint system. The results demonstrate that the proposed method can effectively obtain the optimal matching parameters of occupant restraint system with consideration of the effects of uncertainty and hence ensure the safety of vehicle occupants, having an extensive prospect of engineering application in the field of vehicle safety.

Key words: occupant restraint system, multiobjective uncertainty optimization, local refining, approximate model