基于隐马尔可夫的航空发动机视情维修研究  

Research on Condition Based Maintenance for Aero-engine Using Hidden Markov

作　　者：赵飞[1] 王华伟[1] 周唯杰[1] 

机构地区：[1]南京航空航天大学民航学院,江苏南京210016

出　　处：《航空计算技术》2010年第5期15-19,共5页Aeronautical Computing Technique

基　　金：国家自然科学基金与民航总局联合资助项目(60879001);国家自然科学基金重点基金资助项目(60939003)

摘　　要：针对传统航空发动机维修方式存在的"维修过剩"或"维修不足"的弊端,研究了航空发动机的视情维修决策模型。采用隐马尔可夫过程对航空发动机劣化过程建模,解决了传统马尔可夫过程中故障率均假定服从指数分布的局限,同时考虑了检测过程中的不完全信息情况,更贴近实际工程,相关结果可为视情维修提供辅助决策支持。通过具体示例,验证了方法的有效性和合理性。Aiming at the traditional maintenance system of aero- engine has exposed the abuses such as deficient ma- intenance and excess maintenance, In this paper, we studied the condition based maintenance decision model of aero- en- gine. Hidden markov process was applied to model the degradation process of the aero- engine, which resolved the limita- tion of failure rate was assumed exponentially distributed in traditional markov process. Meanwhile the situation of imper- fect repairs in inspection process was considered, which was much closer to the actual project. And the results can provide assistant decision support for the condition- based maintenance. At the end, the feasibility of the model is verified through a numerical example.

关 键 词：航空发动机 隐马尔可夫 比例风险模型 维修决策 

分 类 号：V37[航空宇航科学与技术—航空宇航推进理论与工程]