考虑部件相关性的复杂设备绩效保证策略设计方法研究

Performance-based warranty is a contractual agreement between customer and manufacturer related to equipment performance. A scientific performance-based warranty policy induces the manufacturer to implement optimal maintenance strategy through reasonable incentive mechanism, making the field performance of equipment beyond the benchmark requirement and achieving the effective cultivation of the manufacturer. This project focuses on the study of key methods of performance-based warranty policy design for complex equipment considering the dependence between components. Firstly, on the basis of condition-monitoring data, the equipment reliability model is developed by incorporating competing failure mechanism, stochastic dependence between components and internal and external influencing factors. Then the remaining useful life (RUL) of the equipment is estimated for condition-based maintenance decision-making. Secondly, according to the RUL prognosis of the equipment, the performance-oriented dynamic condition-based maintenance optimization approach is proposed considering the structural, economic as well as resource dependence between components, in order to support the design of performance-based warranty contract. Thirdly, given the condition-based maintenance strategy by the manufacturer, the performance-based warranty contract is designed considering dynamic incentive mechanism, so as to effectively allocate the benefits of each party in the performance-based warranty chain. The research mentioned above is expected to promote the performance-based warranty research, and enhance the competitiveness of manufacturing industry, especially for manufacturers of complex equipment.

绩效保证是顾客与制造商就设备运行绩效订立的契约性协议。科学的绩效保证策略通过合理的激励机制引导制造商实行最佳的维修策略，尽可能使设备实际绩效凌驾于基准绩效，实现顾客对制造商的有效培育。本课题拟考虑部件相关性，研究复杂设备绩效保证策略设计的关键方法，主要包括：（1）以设备状态监测数据为基础，纳入部件竞争失效机制、随机相关性和内外生因素，构建复杂设备可靠性分析模型，研究设备剩余有效寿命预测方法，为视情维修策略优化提供支持；（2）根据设备剩余有效寿命预知结果，考虑部件结构、经济和资源相关性，研究面向绩效保证的动态视情维修决策优化方法，为绩效保证契约设计提供支持；（3）根据设备视情维修策略，研究考虑动态激励的绩效保证契约设计方法，实现绩效保证链上顾客与制造商利益的合理分配。上述研究对推动复杂设备绩效保证研究有重要理论意义，同时对提高我国制造企业，尤其是复杂设备制造企业的市场竞争力有重要应用价值。

现代复杂设备由多个相互影响的子系统构成，各子系统包含多个功能各异的部件，其失效表现为很强的随机性和关联性，对制造商提升复杂设备的质量和可靠性、健全和完善维修服务提出了更高的要求。绩效保证是顾客与设备制造商就设备运行绩效订立的契约性协议，科学合理的绩效保证策略使顾客通过激励引导制造商主动实行有效的维修策略，尽可能使设备实际绩效凌驾于基准绩效，实现顾客对制造商的有效培育。. 本课题在继承国内外学者对绩效保证相关研究成果的基础上，结合复杂设备维修管理现状，研究了复杂设备绩效保证策略设计的方法。主要研究内容包括：（1）复杂设备可靠性建模与剩余有效寿命预测。量化部件异质性和环境应力等因素对退化失效和冲击失效过程的影响，研究自然退化失效和外部冲击失效的相互作用规律，构建部件竞争失效模型；分析部件的随机相关性，建立基于Copula方法的设备可靠性模型，结合设备当前健康状态，研究设备剩余有效寿命的预测方法。（2）面向绩效保证的复杂设备动态视情维修策略制定。研究部件及群组的维修等级测度方法，识别影响设备可靠性的关键部件和群组；考虑部件结构相关性和经济相关性，建立不同维修情境下设备绩效和成本收益模型；结合设备剩余有效寿命预测结果，研究设备视情维修策略的动态优化方法，获取设备最优动态视情维修策略。(3)考虑激励的复杂设备绩效保证契约设计。研究激励因素对设备制造商视情维修策略的激励效应，建立绩效保证激励模型；考虑主体参与约束和激励相容约束，构建以顾客为委托人、设备制造商为代理人的主从递阶决策优化模型；获取设备最优绩效保证契约，实现顾客和复杂设备制造商利益的帕累托改善。. 上述研究对丰富复杂设备绩效保证管理的相关理论和内容、拓展绩效保证建模与决策的研究有重要理论意义，对指导和服务于复杂设备制造企业的绩效保证实践有重要工程应用价值。