Mechatronic - Advanced Computational Intelligence

Abstract

This paper investigates two different strategies to control nonlinear mechatronic systems. The first method is applicable for a plant with known linear part and unknown static nonlinearity. By an intelligent observer the nonlinear characteristic is identified, such that the result of the identification can be used to improve the performance of the controller. The second approach presented is non-identifier-based. Hence, no identification is necessary if the system possesses some structural attributes. A time-varying control law is employed, which reacts on the present behaviour of the error signal instantaneously. The appealing feature is that bounds for the control error can be fixed a priori.