DEVELOPMENTS IN CONTROL TECHNIQUES FOR NONLINEAR SERVOMECHANISMS

Abstract

This thesis presents new developments in several different control techniques to enhance the performance of control systems applicable to nonlinear servomechanisms. In such servomechanisms, the operating conditions and performance requirements typically vary with different applications. The techniques developed are intended to achieve improved control of the nonlinear servomechanisms and to provide a range of solutions to address the various different problem situations. Five main techniques are developed. The first technique provides a framework for robust neural network-based control of nonlinear servomechanisms. An approach of parallel adaptive neural network-based control is developed as the second major technique. In the third technique, the application and extension of suitable strategies for integrating neural network and fuzzy systems methodologies to enhance a basic fixed neural network-based nonlinear control strategy with the property of self-organization is investigated. A controller with fast load adaptation is developed in the fourth technique, and finally, in the fifth technique. improved learning controllers for nonlinear servomechanism applications involving repetitive operations are developed. In all these developments, the analytical properties of the overall system are described, and the results are verified with experiments and/or simulations.