Abstract: As a crucial influential factor of vehicle quality,the quality of outsourcing parts in the automotive industry not only plays a significant role on vehicles performance and customer experience,but also influences warranty cost,brand reputation and sales amount in next lifecycle.To improve the economics of quality for self-owned brand automobile industry from warranty period perspective,a multi objective mixed integer nonlinear programming model was formulated to derive the optimal supplier portfolio of outsourcing parts by developing a hybrid genetic-based algorithm.The relative importance of key part was calculated by the improved Risk Priority Number (RPN) value concerning the nonlinear characteristics of  Failure Mode Effect Analysis (FMEA) ingredients,and the Taguchi method was employed to reflect hidden quality loss caused by customer complaints.The four items including total quality related cost,system reliability,delivery time and customer complaints were highlighted in the nonlinear multi objective programming model.The multi-attribute utility theory and combined weighting technique was used to integrate the sub-objective functions.To improve the efficiency of the genetic operations,a Hybrid Adaptive Genetic Algorithm (HAGA) was designed by integrating local search strategy to deal with the programming model.The numerical study demonstrated the effectiveness of the model and the advantages of the proposed algorithm,and the results could provide guidance to improve the quality of economics for automotive industries on outsourcing part procurement and supplier portfolio selection from total lifecycle perspective.

Key words: lifecycle, economics of quality, total quality related cost, automotive outsourcing, supplier, portfolio selection, nonlinear multi-objective programming, hybrid adaptive genetic algorithms

摘要： 作为影响整车质量的主要因素,外购件质量的表现不但影响整车在役性能和消费者体验,而且显著影响主机厂三包费用、企业品牌和销量等质量经济性因素。为提升自主品牌汽车的质量经济性,提出基于混合自适应遗传算法的的多目标非线性整数规划模型,预防性优选最佳外购件供应商组合。分析了故障模式及影响分析要素的非线性特性,基于改进风险系数值确定外购件的相对重要度,借助田口损失函数量化由顾客抱怨产生的隐形质量损失。以总质量相关成本、产品系统可靠性、交付期和顾客抱怨数为优化目标,采用多属性效用理论和线性加权方法处理多目标组合优选模型。为提升算法求解效率,引入局部搜索策略求解了非线性多目标规划模型,设计了一种混合自适应遗传算法。通过算例验证了模型的有效性和算法的优越性,为主机厂面向质量经济性视角的关键外购件组合优选提供了借鉴与指导。

关键词: 全生命周期, 质量经济性, 总质量相关成本, 汽车外购件, 供应商, 组合优选, 非线性多目标规划, 混合自适应遗传算法