基于RBF神经网络的非线性系统故障诊断

来源期刊：中南大学学报(自然科学版)2003年第4期

论文作者：贾明兴 陆宁云 王福利

文章页码：455 - 458

关键词：故障诊断;神经网络逼近器;鲁棒性;灵敏度

Key words：fault diagnosis; neural network approximator; robustness; sensitivity

摘    要：针对一类含模型不确定性的非线性系统,提出了具有强鲁棒性和高灵敏度的在线故障检测与诊断方法.其中,系统只有输入、输出可检测,故障是关于输入和状态的非线性函数.非线性在线估计器用于估计系统不确定部分,同时监视系统是否发生故障,估计故障的大小.仿真结果表明,故障诊断算法稳定.

Abstract: A new on-line fault detection and diagnosis method which processes powerful robustness and high sensitivity for a class of nonlinear system with model uncertainty is proposed. The only measurable variables are the inputs and outputs of the system. The faults are assumed to be functions of the inputs and the states of the system. A nonlinear on-line approximator is utilized to estimate uncertain parts in the system, simultaneously, to monitor the faults and estimate the fault value. The diagnosis algorithm is stable. Finally, a simulation example is presented to illustrate the effectiveness of the method.