【摘要】 在汽车有限元模型基础上建立了乘员约束系统模型，进行了100%正面刚性壁碰撞仿真分析。结合碰撞分析中假人伤害值，采用灵敏度分析筛选出对约束系统影响较大的参数作为设计变量，构建了关于加权伤害指标的Kriging代理模型，运用多岛遗传算法进行参数优化。结果表明，加权伤害指标降低了38.92%，达到设计目标。更多还原

【Abstract】 A passenger restraint system model was built based on the finite element model of the car,and a 100% frontal rigid wall collision simulation analysis was conducted. Incorporating the dummy injury values from the collision analysis, sensitivity analysis was employed to identify parameters that have a significant impact on the constraint system as design variables. Subsequently, a Kriging surrogate model for weighted injury indicators was constructed, followed by parameter optimization using a multi-island genetic algorithm. The results show that the weighted injury index decreases by 38.92%,achieving the design goal.更多还原