Output Feedback Controller Design For Complex Systems

At present, people often use state feedback control law to study robust control, but in many practical problems, the system state often cannot be measured directly, it is difficult to use state feedback control law to control the system. Sometimes, even if the state can be measured directly, but, considering the cost of implementing the control and reliability of the system and other factors, the state feedback control cannot achieve acceptable effect. If the output feedback law can achieve the performance requirements of the closed-loop system, then it can be selected with priority. Therefore, the output feedback stabilization of uncertain systems and controller design has important theoretical and practical value.This paper was based on robust control, H-inf control, output feedback control theory. It is based on Lyapunov stability theory and Linear Matrix Inequality (LMI) methods. For uncertain time-delay systems with norm bounded uncertain parameters, the paper studied the output feedback controller and H-inf output feedback controller design methods.First, the paper designed the output feedback controller for the linear uncertain time-delay systems, studied the system with only state time-delay, and designed the static and dynamic output feedback controller for the system with control time-delay and state time-delay. Second, as for the system influenced much by the time-delay, considering the control input with time-delay, this paper discussed H-inf static output feedback control for linear time-delay system with memory, H-inf dynamic feedback control for linear time-delay system with memory and H-inf dynamic output feedback control for dependent time-delay linear system. Third, beginning with the output feedback controller for linear neutral time-delay system, the paper studied a class of neutral time-delay system with nonlinear uncertainty and proposed a new output feedback controller design based on the state-observer which guaranteed that the closed-loop system was of approaching stabilization. The approach of designing the observer and the controller was dependent on the existence of two solutions of LMI. This was simple and it could avoid the trouble of choosing the parameters. The controller parameters were worked out by means of LMI toolbox in MATLAB. Simulation of the actual system was conducted on the basis of the SIMULINK toolbox in Matlab, the results of which proved that the new controller design method could achieve better control effect and was more robust and stable. Finally, the paper discussed the output feedback guaranteed-cost controller design for uncertain discrete systems.