具有Markov跳变参数的广义系统鲁棒故障检测

针对一类广义不确定时滞Markov跳变系统,应用鲁棒控制理论分析重构的增广系统,获取了系统的故障检测滤波器.利用构造的随机Lyapunov-Krasovskii函数和线性矩阵不等式,证明并给出了故障检测滤波器有解的充分条件和优化设计方法,同时使得残差体现其对扰动的抑制作用和对故障的灵敏性.所设计的滤波器使系统具有随机稳定性,抑制干扰能力强,满足所给的范数指标.仿真算例对结果进行了验证,在故障信息发生时,故障检测滤波器可以很灵敏地检测出故障,并能有效地抑制未知扰动对残差的影响在给定的范围内.

ROBUST FAULT DETECTION FOR SINGULAR SYSTEMS WITH MARKOV JUMP PARAMETERS

The fault detection filter is designed by robust control theory for a class of uncertain singular time-delay system with Markov jump parameters. By using the constructed stochastic Lyapunov-Krasovskii functional and linear matrix inequalities techniques, a sufficient condition is provided to solve the design of fault detection filter, and the design problem is described as an optimal one. Simultaneously, the residual error is selected separately to reflect the restraint of disturbance and the sensitivity of faults. The presented fault detection filter makes the systems have stochastically stability and better ability of restraining disturbances, and satisfies the given norm index. Simulation examples illustrate the results. As the fault signals happen, the proposed filter can detect them sensitively, and the influence of unknown disturbance on residual error can be restrained effectively to a given range.