Because of the rapid development of computer network technology, and the falling price of computer hardware, networked control systems of automation gradually come into our lives. Remote control and complex system control is the most pressing computer technology needs of the current human beings. Through the analysis of these factors, doing the comprehensive analysis of the networked control systems is a important point of what we need urgently to solve the control theory in the research category.We have used some basic theories, such as the switched system theory, the stochastic control theory, the timedelay system theory, the robust control theory and the linear matrix inequality (LMI), etc., to deeply study about certain linear systems, no rules of uncertain systems and the network control of nonlinear systems.In this paper, first, the problem what we studied is the networked control systems exist timevarying delay, time varying transfer interval and communication constraints simultaneously, based on the above premises, we give a definition to the stability of the networked control systems. We have analysised the steady state performance of the networked control systems based on the condition that the transfer interval and transmission delay are regarded as continuous random variable sequences, as the random agreement seems to be a switching function. This analysis above is based on the random changing parameters discrete time switched linear system models of networked control systems. By adopting the excessive approximate method of dealing with continuous probability distribution in meansquare, we get the methods to guarantee the stability of system, and to guarantee the current stability will not affected by the early results before.We discussed the robust control stability problem of the networked control system with short delay. We introduced the problem of robust control and robust stability. We use a new, linear, random system to describe the networked control systems, at the same time we ues a new state feedback control law to design the stochastic system. Through deeply analysis of the closed loop networked control systems, we obtained its gradual stability conditions and established the quantitative relationship between the maximum delay and delay variation range which is satisfied the gradual stability conditions. By analysis the final simulation results, we can know that this method is feasible.We discussed the the fault detection of the networked control systems with short delay in the case of no packets lost,. As a first step, two basic models of networked control systemsthe timedriven sensor model, as well as the eventdriven controller and actuator model, this paper made a deep analysis and explain in detail, and established a discrete equation with uncertain parameters; For fault detection of the system, we designed a state observer; We can make use of lyapunov stability criterion to get the condition of the stability of the fault detection system. These mentioned above are all validated by the simulation results which shows that this method can be implemented.At the end of this article, we summarizes the key work and the main innovation points, expand the width of them so as to make the research work a solid foundation for the future.
