摘要
在常规基于解析冗余故障诊断技术的基础上,采用具有最佳模拟特性的RBF神经网络对系统进行建模,分析了M距离应用于卫星姿态控制系统故障检测与定位的可行性,应用基于M距离的方法设计故障检测观测器,通过对残差的评估实现故障诊断。仿真结果显示,该方法计算过程简单、实时性好。
Based on the traditional analytical redundancy fault diagnosis technology, the RBF neural network with the best imitating performance is adopted to model the control system. Then the feasibility of detecting and identifying faults for the satellite attitude control system with the Mahalanobis distance is analyzed in detail. At last the fault-detection observers are designed and implemented based on the residual evaluation so as to achieve the fault diagnosis. Simulation result indicates that the diagnosing method proposed in this paper is characterized with light computation burden and good real-time property.
出处
《航天控制》
CSCD
北大核心
2006年第6期61-66,共6页
Aerospace Control
关键词
M距离
故障诊断
姿态控制系统
残差
RBF神经网络
Mahalanobis distance Fault diagnosis Attitude control system Residual RBF neural network
