基于Kriging模型的大客车侧翻安全性多目标优化

Multi-objective Optimization for Bus Rollover Safety Based on Kriging Model

建立某型客车的有限元模型,依据ECE R66法规对客车侧翻安全性进行了多目标优化.根据侧翻仿真结果,选取客车上部结构关键零件截面形状为变量,以最大质心加速度和质量为目标,以侧围立柱与地板间夹角θ为约束条件,并结合最优拉丁方试验设计法生成了高精度的Kriging模型.利用NSGA(non-dominated sorting genetic algorithm)-II算法对大客车侧翻安全性进行了多目标优化并得到了Pareto最优解集.优化结果表明利用截面形状优化法可以在质量增加较小的条件下提高整车的侧翻安全性能.

A finite element model of a bus is established, and according to the ECE R66 regulation, multi-objective optimization is carried out for the bus rollover safety. First, on the basis of the rollover simulation result, the section shapes of the key parts in the upper structure are selected as design variables. The maximum acceleration of the center of gravity and the mass are then taken as objectives, and the angles between side pillars and the floor are selected as constraints. Then, combined with optimal Latin Hypercube, the Kriging models of high accuracy are constructed. Finally, multi-objective optimization for the bus rollover safety is performed with NSGA-II, and the Pareto optimal set is obtained. The conclusion is drawn that the rollover safety can be improved by using section shape optimization when a limited mass is increased.