基于概率盒的乘员约束系统混合可靠性优化设计  

Hybrid reliability-based design optimization of the occupant restraint system based on the probability box

作　　者：王琼[1] WANG Qiong(School of Mechanical Engineering,Hunan Institute of Engineering,Xiangtan 411104)

机构地区：[1]湖南工程学院机械工程学院,湖南湘潭411104

出　　处：《机械设计》2022年第6期51-57,共7页Journal of Machine Design

摘　　要：文中针对不确定参数的概率分布函数存在样本信息缺乏的非精确概率问题,基于参数化概率盒构建乘员约束系统混合可靠性优化设计模型。由于存在区间参数,内层以约束可靠度的区间下界来度量可靠性,从而确保乘员约束系统的安全性设计。采用一种基于渐进移动矢量策略的高效解耦算法,将嵌套优化问题转换为确定性设计优化与混合可靠性分析的序列迭代过程,从而避免内外层嵌套寻优,以实现乘员约束系统混合可靠性优化设计的高效性。工程应用案例结果表明:优化后综合损伤值降低,同时约束可靠度指标均得到满足,经过少数几个迭代步实现了乘员约束系统混合可靠性优化设计。Since the probability distribution functions of the uncertain parameters suffer from inexact probability with inadequate sample information, in this article, the hybrid reliability-based design optimization(HRBDO) model of the occupant restraint system is built based on the parameterized probability box. Given the interval parameters, reliability in the inner layer is qualified by limiting the lower bound of reliability interval, thus ensuring that the occupant restraint system is designed in a safe manner. The efficient shifting vector-based decoupling algorithm is adopted, which converts nested optimization into deterministic optimization and sequential iterative process of the hybrid-reliability analysis, therefore avoiding nested optimization between the inner and outer layers, in order to achieve efficient reliability-based design optimization of the occupant restraint system. The results of engineering cases indicate that after optimization, the comprehensive injure value reduces and all the constraints of the reliability indicators are met. As a result, hybrid reliability-based design optimization of the occupant restraint system has been realized in several iterative steps.

关 键 词：混合可靠性优化设计 乘员约束系统 非精确概率 参数化概率盒 

分 类 号：U462.3[机械工程—车辆工程]