摘要
以卫星姿控系统实时仿真信号为诊断依据,设计故障检测Elman神经网络及故障判决,实现系统正常与非正常状态的区分并获取故障发生时刻。提出了基于改进梯度更新策略的故障隔离Elman神经网络方法,对故障时刻点之后时域信号进行故障模式匹配,进一步实现系统故障隔离。运用某卫星姿态控制系统进行在线故障诊断试验的结果表明,本文方法具有较好的实时有效性、输出耦合诊断性能、时域信号诊断泛化性和网络收敛性。
An online fault detecting and isolating (FDI) method was proposed for fault diagnosis of Infrared Earth Sensor.The Elman neural network was employed due to its capability of processing time-varying signals in real time.For the sake of simplicity,two Elman neural networks were developed for fault detecting (FD) and fault isolating (FI) respectively.For FD,an FD Elman neural network and corresponding logic judgment were designed to identify normal and faulty states; the output of FD was the moment when a fault occurred.For FI,a novel gradient updating strategy was introduced in FI Elman neural network which does faulty pattern matching and conducts FI.Simulation results demonstrate that the FDI strategy is real time,convergent available for output coupling,general with time-varying signals.The proposed FDI can avoid modeling,so it is suitable for online FDI of satellite attitude control system (SACS).
出处
《振动．测试与诊断》
EI
CSCD
北大核心
2010年第5期504-509,共6页
Journal of Vibration,Measurement & Diagnosis
基金
国家高技术研究发展计划("八六三"计划)资助项目(编号:2007AA04Z438)
