Analysis And Control Of Networked Control Systems With Delay And Packet Dropout

Real-time feedback control system in which control loops are closed through a serial network is called networked control system(NCS). With the development of control and network technology, NCS has now found applications in a number of new areas, such as the Internet of Things, by overcoming many constrains, such as network-induced delay, packet dropout, and noise disturbance. Therefore, the research on NCS creates great academic value and considerable economic benefits.This thesis is concerned with the modeling, stability analysis, and controller design for NCS by taking all the effects of of network-induced delay, packet dropout, node-driven pattern, and external disturbance into considerations. The proposed analysis and synthesis are mainly based on uncertain system approaches, linear matrix inequalities(LMI) techniques, markov chain jump systems, and robust control theories. The main work and contributions are as follows:1. Robust H?controller design problem is studied for NCS with time-varying delay and noise disturbance. By considering state feedback control method, NCS is modeled as a discrete-time linear uncertain system. Meanwhile, uncertainty of networked-induced delay is transformed to uncertainties of system models, and noises are considered to exist in both the sensor-to-controller channel and controller-to-actuator channel and have different characteristics. Then, a sufficient condition for the asymptotic stability are established, and the condition ensures that the closed-loop NCS achieves an H?noise attenuation performance ?. Furthermore, design methods for the robust controller are also presented.2. Stabilization is studied for NCS with short time-varying delay and asynchronous packet dropout. Time-varying delay and asynchronous packet dropout exist in both the two channels of the network. Meanwhile, random delay is described by uncertain system parameter matrix. Then, NCS is described by asynchronous dynamics systems with four subsystems and with rate constrains on events. Sufficient conditions for the exponential stability of the closed-loop NCS is presented in terms of LMIs. Further, design methods for the state feedback controller are also presented.3. The robust?H controller design problem is studied for NCS with short time-varying delay, markovian packet dropout and external disturbance. These three parameters exist in two channels of the network. Random delay is described by uncertain system parameter matrix. Then, NCS is modeled as a markov jump linear system with four subsystems. With the changing network conditions, the system jump among four subsystems above. Besides, sufficient conditions for the NCS to be stochastically stable and ensure an H?performance level ? are presented in terms of LMIs. Finally, the design methods for the dynamic output feedback robust controller are also presented.Illustrative examples are provided to show the correctness and effectiveness of the proposed methods. Moreover, the proposed method is applied to the internet transmission system of attendance machine in Internet of Things, and stability of the system has been greatly improved.