Abstract:

Aiming at the complexity characteristics of automobile safety optimization considering the effects of stamping, a compensation response surface model was created by using optimal Latin square design through the small sample data spanning the traditional model and high accuracy model. And then a new Kriging response surface model was established on the basis of new test data which was created through the traditional model and compensation response surface model. A multi-objective particle swarm optimization algorithm was used to solve and optimize the response surface model.  The method was applied to vehicle frontal impact and lightweight multi-objective optimization design. The results show that this method can ensure the response surface precision, rapid convergence to the optimal solution.

Key words: automobile safety, sheet metal forming, variable complexity, approximate model, multi-object optimization

摘要：

针对汽车安全性优化过程中考虑冲压效应的计算复杂性特点，采用最优拉丁方实验设计，利用较少样本点数据在传统模型和高精度模型间建立一个差值补偿响应面模型，再通过传统模型和差值补偿响应面模型构造新的实验样本点，在此基础上建立Kriging响应面模型，应用多目标粒子群优化算法对此响应面模型进行优化求解，进行了整车正面碰撞和轻量化多目标优化设计。结果表明，该方法能够在保证响应面精度的前提下，快速收敛于优化解。

关键词: 汽车安全性, 冲压, 变复杂度, 近似模型, 多目标优化