具有非线性扰动多时滞系统的鲁棒故障检测  

Robust fault detection for multiple time-delay systems with (nonlinear) perturbations

作　　者：张登峰[1] 孙金生[1] 王执铨[1] 胡寿松[2] 

机构地区：[1]南京理工大学自动化系,江苏南京210094 [2]南京航空航天大学自动化学院,江苏南京210016

出　　处：《控制与决策》2004年第3期358-360,共3页Control and Decision

基　　金：国家自然科学基金重点项目(60234010);国防科技行业重点预研基金资助项目(40404070102).

摘　　要：基于状态观测器的方法研究了一类具有非线性扰动的多重状态时滞系统的鲁棒故障检测问题.应用RBF神经网络逼近系统的非线性扰动,采用线性矩阵不等式(LMI)给出了与时滞上界相关的增益阵设计方法,并利用Lyapunov函数和一致有界引理证明了故障检测残差信号的一致有界稳定条件和对非线性扰动的鲁棒性.仿真示例说明了该方法的有效性.A robust RBF ANN-observer-based fault detection scheme is proposed for a class of multiple state time-delay systems with nonlinear perturbations. An LMI approach, connecting with upper bound of time-delays, is presented to design the observer gain-matrix. The uniform bounded stable condition of the residual and its robustness to nonlinear perturbations are proved by using Lyapunov function and uniform bounded lemma. Numerical simulation demonstrates the effectiveness of the proposed scheme.

关 键 词：故障检测 状态观测器 时滞系统 非线性扰动 线性矩阵不等式 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]