【摘要】 复杂系统故障诊断中的一些关键技术问题一直制约诊断技术的发展.本文对其中的两类典型的关键技术(性能衰退预测技术、故障定位技术)进行了介绍和分析.简要总结分析了复杂系统故障诊断当前的研究现状.提出了复杂系统故障诊断中的两类关键技术(性能衰退预测、故障定位),并给出了适合这两类关键技术的具体方法.结合作者所在课题组的研究,介绍了基于àTrous算法的小波递归预测的小卫星电源系统性能衰退预测技术、基于双级径向基函数(RBF)神经网络的液压伺服系统故障定位技术.研究结果证明了所提出的方法对于这两类典型对象故障诊断的有效性.更多还原

【Abstract】 Some key technical problems have always been the bottlenecks constraining the development of complex system fault diagnosis technology. Performance degradation prediction and fault location,two of the typical and key problems,were discussed and analyzed in this paper.Research status of complex system fault diagnosis was briefly reviewed.Performance degradation prediction and fault location,together with the corresponding solution approaches,were proposed as two key techniques.Two relative study cases on fault diagnosis of complex system were presented,including à Trous wavelet recursion prediction based performance degradation prediction of small satellite power system,hierarchical radius basis function(RBF) neural network based fault location of hydraulic servo system.It was found that the proposed approaches were effective and pragmatic for the two types of systems.更多还原