Robustly Networked Control Systems Subject To Network-induced Delay And Packet Dropout

With the development of communication network technology, there is an increasing tendency to apply communication network to control systems where a lot of sensors, controllers and actuators exist. The control systems wherein control loops are closed through communication network are called networked control systems. Compared with the traditional control systems, networked control systems have advantages of less wire and volume, low implementation cost, easy maintenance, high flexibility and easy tele-operation and so on. So networked control systems can be observed in many areas such as modern manufacturing plants, cars, aerial vehicles, tele-operation etc. But, the communication network in the control loop brings about the problems of network-induced delay and packet dropout. It is difficult to model the controlled plant exactly in real systems. Many scholars have researched robust control of networked systems in order to make networked control systems less sensitive to the uncertainty and external disturbance. However, there still exist some problems to be solved. The following problems are studied in this dissertation:1. The stabilization of networked control systems with random network-induced delay and asynchronous packet dropout is considered. The network-induced delay is divided into the constant average delay and the uncertain delay and transformed into the coefficients uncertainty of state space equation. Networked control systems are modeled as asynchronous dynamical systems with two events constraints. According to the theory of asynchronous dynamical systems, the stabilization of networked control systems is studied. The sufficient conditions of exponential stability of state feedback and dynamical output feedback networked control systems are given respectively. The controllers can be obtained by the conditions.2. The networked controlled plant is often disturbed. The robust H_∞control of networked control systems with random network-induced delay and packet dropout governed by Markov chain is studied. The networked control systems are modeled as markovian jump linear systems with two modes. The sufficient condition of state feedback robust H_∞controller is given in terms of linear matrix inequalities. When the states of the plant are difficult to measure, the sufficient condition of output feedback robust H_∞controller is presented based on linear matrix inequalities. The designed controllers can stabilize the networked control systems and attenuate the disturbance to a prescribed level. 3. When the controlled plant is a system with parametric uncertainties, the linear quadratic optimal control is difficult to implement. The guaranteed cost control of networked control systems with random network-induced delay and packet dropout governed by Markov chain is addressed. The networked control systems are modeled as markovian jump linear systems with two modes. The sufficient condition of guaranteed cost controller is given in terms of linear matrix inequalities. The guaranteed cost controller makes not only the networked control systems stochastic stability but also the performance index less than an upper bound.4. When the sensor data and control data of networked control systems are lumped into multiple packets to transmit, the network-induced delays of the packets are different from each other. Based on the relationship between the outputs and inputs of the plant and the controller, the networked control systems with multiple packets transmission are modeled as system with multiple delays. A new delay dependent stability condition of networked control systems is obtained by taking the relationship between the delay upper bounds into account in Lyapunov functional and estimating the upper bound of the derivative of Lyapunov functional without ignoring the additional useful terms. Compared with the existing results, the proposed condition is less conservative.