摘要
针对定性模型诊断方法在处理多故障模式复杂系统时计算量大,不能满足航天器自主诊断实时性要求的问题,提出了新的诊断方法。为了减小冲突识别的实时计算规模,引入冲突识别的分离策略,提出了解析冗余计算方法。扩展了候选产生算法,先识别可能的异常部件,再通过冲突环境对异常部件的故障类型进行确定,在实现多故障模式系统故障诊断的同时,缩小了匹配空间。以某卫星一次电源系统为例,通过与现有算法进行对比,验证了本文的方法能够有效地减小诊断的规模,提高诊断效率,满足航天器自主诊断的需求。
In order to solve the large amount of calculatial problem of the failure diagnosis approach for a multiple failure mode complex system based on qualitative models, a new approach is proposed in this paper. Aiming at reducing the real-time computational scale, a method is proposed for calculating the analytic redundancy when a separation strategy for conflict recognition is adopted. The candidate generation method is extended by recognizing the failure component firstly, and the confirming the failure-mode. This method can reduce the matched space effectively while implementing the failure diagnosis of multiple failure mode system. This approach is applied to a primary electrical power system of some satellite, results of comparison with the existing method show that the approach can reduce the amount of calculation, and improve the diagnosis efficiency.
出处
《宇航学报》
EI
CAS
CSCD
北大核心
2012年第6期698-704,共7页
Journal of Astronautics
基金
国家重点基础研究发展基金(973计划)(2012CB720003)
国防重点学科实验室开放基金(HIT.KLOF.2011.074)
关键词
故障诊断
解析冗余
多故障模式
一次电源系统
Fault diagnosis
Analytic redundancy
Multiple failure mode
Primary electrical power system
