Sliding Mode Reduced Order Control Design And Application For A Class Of Nonlinear Affine System

For practical engineering applications, the majority of non-linear systemincluding robot system, power system, helicopter controlling system, and controllingsystem for industry process, have the mathematical models with affine nonlinear formassociated with massive size. Therefore, it is very instructive to investigate thereduction of the model order for such non-linear systems. The research work of thisdissertation mainly include4aspects, described as follows.1. The research content and state of the art of the model order reduction wasreviewed after referring to related literature, and the strength and weakness of eachmodel order reduction approach was also analyzed.2. In this dissertation, we study a type of nonlinear affine system controllerdesign problems, and a new controller design scheme, i.e. the reduced order of slidingmode, has been proposed in light of sliding mode control theory. Firstly, for a type ofhigh order affine nonlinear system, the appropriate diffeomorphism transformationfunctions and sliding manifold are built in order to reduce the initial high order systemto low order system by using variable structure control theory repetitively as well asconstructing variable structure control rule. Secondly, the control input of the initialsystem is obtained reversely using the mapping relationship between the control inputof the current system and that of the last system. The effectiveness and feasibility ofthis strategy has been demonstrated through the simulation.3. This sliding mode order reduction controller design method has been appliedto the multiply input CE150helicopter model with cross coupling effect for slidingmodel mode order reduction and controller design. In addition, the suppression of theexternal system interference and elimination of the chattering are analyzed. Finally,The numerical simulation performed by MATLAB has verified the effectiveness ofthe proposed method that also exhibits a certain robustness.4. The conclusion and the future research directions are given at the end ofdissertation.Compared with other model order reduction method, the innovation of thesliding model order reduction approach proposed in this dissertation is shown in thefollowing three aspects: first of all, this approach fits model order reduction for nonlinear control systems; furthermore, this method does not lose any information oforiginal system compared with traditional order reduction method; finally, thegenerality of this sliding manifold design is found, indicating a widespreadapplication.