一类非线性系统的渐变故障诊断

Slow-Varying Fault Diagnosis for a Class of Nonlinear Systems

研究非线性观测器故障诊断优化问题,针对一类状态不能测量的非线性不确定动态系统,提出了一种用RBF神经网络逼近渐变故障的诊断方法。设计非线性观测器来估计系统的状态,利用Lyapunov稳定性定理给出了RBF神经网络的权值、中心、宽度的更新调节律,通过在线调整RBF神经网络的权值、中心、宽度监测和估计系统中发生的非线性故障,实现了状态不能测量的非线性系统渐变故障诊断。最后,仿真例子证明了方法的有效性。

A fault diagnosis method based on RBF neural networks on-line approximation structure for uncertain nonlinear system was presented in this paper. A nonlinear state observer was designed to estimate system states that were inputs to the neural networks. An adaptive RBF neural network with on-line updated weight, centre and the width vector was used to approximate the slowing-varying faults. The adaptive parameter updating laws were derived using the Lyapunov stability theory. The slowing-varying fault diagnoses can be realized for the nonlinear system, in which all the states are not measured. Simulation examples were used to illustrate the effectiveness of the proposed method.