Operator-based Nonlinear Robust Controller Design And Applications

In real applications, the control systems are always of nonlinearproperties. Besides, there always exist unknown modeling errors andexternal disturbances in real environment, so the obtained models alwayscontain uncertainties consisting of modeling errors and disturbances.Therefore, how to guarantee robustness in designing control system fornonlinear systems in presence of uncertainties is still a challenging issue. Asa result, in this thesis, an operator-based robust tracking control fornonlinear system with unknown uncertainties is investigated by using robustright coprime factorization approach.Firstly, for the nonlinear feedback control system which has beenensured as robust stabilizing system, a PI tracking controller is designed toguarantee the tracking characteristics. However, because of the exitingunknown uncertainties, the perfect tracking condition can not be obtained byusing this method. To solve this problem, combined with robust rightcoprime factorization approach, a sliding mode tracking controller based onthe improved exponential trending law is designed. An advantage of slidingmode control is fast response, in other words, it could shorten the responsetime compared with PI control. However, the disadvantage is high-speedswitched characteristic resulting in chattering. The effectiveness of theproposed method is confirmed by the simulation results of two-link rigidrobot arm. Compared with PI control, the tracking response times ofposition and speed are reduced to60%.Besides，to many control methods for nonlinear systems, it is very hardto eliminate or restrain the effects of uncertainties. In the present control system design, the effect of the disturbances and modeling errors on thesystem performances are considered as uncertainties of system dynamics.Considering the uncertainties, a robustly stable control based on anoperator-based observer is investigated, and the conditions of robust stabilityand perfect tracking are obtained.Robot arm is a typical nonlinear system with strong dynamicalcoupling effects，and is sensitive to unknown modeling errors and externaldisturbances. The robustly stable control method based on an operator-basedobserver is applied to control design of two-link rigid robot arm, theeffectiveness of the proposed method is confirmed by the simulation results.Compared with the former method, the error of trajectory of the robot arm isreduced to75%, realizing perfect tracking control.