摘要
为了获得汽车乘员约束系统理论最优解,针对优化中设计参数多且变化范围大使得设计工作量大且难的问题,采用敏感性分析的结果,提出了代理模型设计域更新的优化方法。通过优化的拉丁方在设计域中抽取少量样本点构造Kriging代理模型,计算得到初始的最优解,再通过敏感性分析的结果快速确定下一轮迭代设计域,从而在优化过程中不断构建代理模型,设计域不断更新,直到优化结果满足收敛条件。通过4次迭代使得乘员综合损伤值WIC相比初始设计值降低了13.83%,优化效果明显。
The optimization design of the occupant restraint system involves huge computer, and the optimal solution in theory is also difficult to obtain. A new method is proposed that combining the sensitivity analysis result with metamodel to update the de- sign space. By using the optimal Latin hypercube method to extract sample points in the design space and constructing the Kriging metamodel and calculating the initial optimum solution. Using the initial optimum solution and the sensitivity analysis resuh to de- termine the design space of next iteration quickly,in the process of optimization, the metamodel and the searching space are con- tinually updated, until the optimization iterations terminate with the convergence condition satisfied. By putting this method to op- timization design of a vehicle occupant restraint system, after four iterations, compared with the initial design values, the compre- hesive damage value WIC reduced by 13.83% ,the effects of optimization is obvious.
出处
《现代制造工程》
CSCD
北大核心
2015年第9期64-69,共6页
Modern Manufacturing Engineering
基金
国家自然科学基金资助项目(51275164)
关键词
敏感性分析
代理模型
设计域
乘员约束系统
sensitivity analysis
metamodel
the design domain
the occupant restraint system
