Stabilization Of Nonlinear System By Sampled-data Output Feedback

The main purpose of this thesis is to study the problem of stabilization ofnonlinear system by sampled-data output feedback. Sampled-data controller isdifferent from the traditional controller. It uses sampled useful information to designthe controller and output the controller through zero-hold to system. Therefore, itcan only use information of fixed time points and output the controller in specialform, then it is difficult to design the sampled-data controller to stabilize the system.One goal of this thesis is to design the sampled-data output feedback controllerto stabilize a class of non-minimum phase nonlinear system. As known, fornon-minimum phase system, the zero-dynamic part is unstable and it is difficult tostabilize. Here the thesis will design the high-gain observer at first, then design thecontroller for a class of lower-triangular system and search condition that thecontroller can stabilize the system. Also the similar method is used to stabilize theupper-triangular system, but controller cannot be designed at a time aslower-triangular and two times transformation should be used.On the other hand, the thesis wants to extend the result of sampled-data outputfeedback controller to a bigger class of nonlinear system. Based on the purpose anda result that the system can be semi-globally asymptotically stabilized by outputfeedback if the system has a structure of lower-triangular, a sampled-data outputfeedback controller is designed to semi-globally asymptotically stabilize the systemunder condition that sampling period is small enough and system can be locallyexponentially stabilize and globally asymptotically stabilized by state feedback.All above results are based on the conclusions that if the system can beexponentially stabilized by output feedback and the sampling period is small enough,then the system mostly can be stabilized by sampled-data output feedback.