The Reliability Of The Network Control System Containing Random Failure Control And Filter Design

The NCS(Networked Control System) is the feedback control loop through the communication network. It has the advantages of less wiring, lower maintenance and diagnostic cost as well as greater the flexibility of the system and so on. But, at the same time, the NCS brings some problems, such as the networked delay, packet loss and networked noise. Based on these problems existing in the NCS and random fault in the actuator or sensor, this paper deals with modeling the system, reliability control and filter design of the NCS. The main work of this paper is as follows.Firstly, the paper discusses the reliable control problem of the general linear time-delay system. Considering the random fault in the actuator and using the Lyapunov-Krasovskii function method, robust H∞controller gain is designed with the Matlab solving the LMI(Linear Matrix Inequalities). It should be noticed that the uncertainties of the fault rate estimation are considered in this paper. A numerical example is given to demonstrate the efficiency of the proposed approach.Secondly, this paper researches the robust fault-tolerant control for the NCS. Considering the random fault in the actuator for the NCS with packet loss and networked delay, Bernoulli probability distribution is used to model the fault status of the actuator. It should be noticed that not only the fault is mentioned above, but also the uncertainties of the fault rate estimation are considered. Using the Lyapunov-Krasovskii function method, sufficient conditions for the robust mean square stability are obtained, and the robust H∞controller gain is designed based on the LMI. A numerical example is given to demonstrate the efficiency of the proposed approach.Finally, this paper proposes robust H∞filter design for the NCS with random networked delay, packet loss and random fault. Considering the random fault in the sensor, Bernoulli probability distribution is used to model the fault status of the sensor. It should be noticed that not only the fault is mentioned above, but also the uncertainties of the fault rate estimation are considered. Using the Lyapunov-Krasovskii function method and exponential stability method, sufficient conditions for the robust mean square stability are obtained, and the robust H∞filter is designed. The example simulation indicates the effectiveness of the presented method.