基于可诊断性约束的测点优化配置研究  

Fault Diagnosability Constraint Based Optimistic Sensor Placement

作　　者：刘文静[1,2] 刘成瑞[1,2] 王南华[1,2] 

机构地区：[1]北京控制工程研究所,北京100190 [2]空间智能控制技术重点实验室,北京100190

出　　处：《空间控制技术与应用》2011年第2期1-5,25,共6页Aerospace Control and Application

基　　金：国家自然科学基金资助项目(61004073);CAST创新基金课题(CAST2011-06)

摘　　要：为提高卫星的安全性和可靠性,适应在设计阶段对故障诊断进行考虑的需求,以动量轮为例,将故障可诊断性作为优化目标,提出基于DM分解技术的测点优化配置方法,并给出故障可诊断性评价的相关定义.分析动量轮模型和故障,建立动量轮的结构模型;采用DM技术对动量轮的结构模型进行分解,并分析动量轮模型的解析冗余关系;在对各故障的可检测性和可分离性进行分析的基础上,根据结构模型的偏序关系,获得使故障具有可检测性和可分离性的最小传感器集合;给出故障可诊断性评价的定义,并针对两种传感器配置情况,进行可诊断性分析.本文所提方法为设计阶段开展故障诊断的研究提供了借鉴和参考.In order to raise the satellite＇s safety and reliability,as well as to meet the requirements for fault diagnosis considered at the design stage,taking a momentum wheel as example,an optimistic sensor placement method based on DM decomposition is proposed by considering the fault diagnosability as the optimized object and the related definitions are given.Based on the analysis of the momentum wheel model and the related faults,the structural model is built,and then the DM decomposition is used to find the analytical redundancy relation.According to the partial order relation in the structural model,the minimal sensor sets which make all the faults have the maximum detectability and fault isolability are obtained.For the two cases of sensor placement,some related definitions are given,and the fault diagnosabilitys are evaluated.The proposed methods give a valuable reference to the fault diagnosis research.

关 键 词：传感器优化配置 DM分解 可检测性 可分离性 动量轮 

分 类 号：TP306.3[自动化与计算机技术—计算机系统结构]