摘要： 针对具有时变采样周期和时延的网络化控制系统(NCS)故障检测问题，通过矩阵Jordan变换，将时变采样周期与时延的不确定性转化为NCS系统结构参数的不确定性，建立离散时间凸多面体不确定系统模型。设计鲁棒故障检测滤波器(RFDF)作为残差发生器，将故障检测问题转化为RFDF的设计问题，通过线性矩阵不等式给出滤波器存在的充分条件。数值仿真结果表明，所设计的RFDF滤波器能较快复现故障信号，对外部干扰和不确定的采样周期及时延具有鲁棒性。

关键词: 网络化控制系统, 时变采样周期, 时延, 凸多面体不确定系统, 鲁棒故障检测滤波器, 线性矩阵不等式

Abstract: Aiming the problem that robust fault detection of Networked Control System(NCS) with time-varying sampling period and time delay, the uncertainties of time-varying sampling period and time delay in NCS are converted into the uncertainties of the parameter matrix by the real Jordan form approximation, and a discrete-time model with parameters uncertainties lying inside a polytypic framework is proposed. The residual generator is constructed and the problem of fault detection can be converted into the design of the Robust Fault Detection Filter(RFDF). The sufficient conditions for existence of the RFDF are established in terms of linear matrix inequalities. Simulation result shows the filter can repeat the fault signal well and it is robust for the disturbance and the uncertain sampling interval and time delay.

Key words: Networked Control System(NCS), time-varying sampling period, time delay, convex polytopic uncertainty system, Robust Fault Detection Filter(RFDF), Linear Matrix Inequality(LMI)

中图分类号: 

• TP393