Output Feedback Control Of A Class Of Multi-input Multi-output Uncertain Nonlinear Systems

Multi-input multi-output(MIMO)nonlinear systems widely exist in a lot of practi-cal engineering.Since the early of 1970's,the control of MIMO nonlinear systems has been a hot topic to researchers of control community.However,compared with single-input single-output(SISO)nonlinear systems,MIMO nonlinear systems are of strong coupling in the system dynamics,resulting in rapid increased difficulty of controller design.Therefore,there are comparatively few research results about the control of MIMO nonlinear system-s over the last four decades.Considering that the research progress of MIMO nonlinear systems theory has a strong influence on the development of modern industrial science and technology,it is quite of theoretical and practical significance to explore much more effec-tive control methods for MIMO nonlinear systems.In order to promote the industrialization process of modern society,it is urgent to accelerate the research on MIMO nonlinear sys-tems theory.This thesis mainly investigates the output feedback tracking control problem of a class of MIMO uncertain nonlinear systems.And the main contents are listed as follows:(1)This thesis presents an original output feedback control scheme for a class of MI-MO uncertain nonlinear systems.Based on backstepping methods,the error system model is developed gradually,and virtual controllers are proposed.In order to deal with the system uncertainties,neural networks are used to approximate the uncertain functions online.Based on a single-layer neural network,a state feedback controller is presented to make the error system achieve uniformly ultimately bounded results.Applying neural networks into observer design,a new output feedback control law is provided.By input of the above output feedback controller,the closed-loop error system is uniformly ultimately bounded.(2)This thesis provides a robust output feedback control law for a class of MIMO uncer-tain nonlinear systems with input and output delays.With a simple and effective way to define auxiliary error signals,the error system model is developed and thereby an original state feedback control strategy is proposed to make the error system achieve uniformly ultimately bounded results.However,the above state controller cannot be implemented online since it contains "future state".To predict the "future state",a high-gain predictor is developed based on high-gain observers.Replacing the "future state" in the above state feedback controller with the predicted value of the "future state",an output feedback control law is presented and the closed-loop error system is uniformly ultimately bounded.