Robust Control System Design for a Class of Uncertain Nonlinear Systems

Abstract

Robust Control System Design for a Class of Uncertain Nonlinear Systems The work presented in this thesis is an endeavor to combine two nonlinear control frameworks in order to utilize the beneficial attributes of adaptive backstepping and sliding mode control or integral sliding mode control techniques. The adaptive backstepping method is capable of eliminating external disturbances and dynamically estimating the unknown parameters. This technique solves the problem of relative degree which is the main drawback of conventional sliding mode control method. It is also globally effective for the stability of the nonlinear control systems. The sliding mode control is another widely used method due to its robustness against matching uncertainties. The robustness of nonlinear system can be enhanced by adding an integral term with the sliding manifold of sliding mode control method. This attribute is achieved by eliminating the reaching phase of sliding mode control which makes it immune against fast parametric variations that occurs during reaching phase. On the basis of this synergy, new strategies have been proposed which are named as Adaptive Backstepping Higher Order Integral Sliding Mode Control (ABHOISMC), and Adaptive Backstepping Integral Sliding Mode Control (ABISMC), and based on our simulation studies, they are found to be more robust against both matching and mismatching uncertainties. During the implementation of newly proposed techniques, it is not necessary to transform the system into triangular form which is generally needed in the conventional backstepping procedure. The first proposed scheme, the ABISMC, is applied to control the parameters of continuous stirred tank regulator plant with its model expressed mathematically in a non-triangular form. The simulation results establish the efficacy of the proposed scheme. As a second example, the ABISMC is also applied for the tracking of the desired output in a field-controlled direct-current motor. The simulation results are found to be very convincing. Finally, the second proposal, the ABHOISMC, is applied on a theoretical plant expressed in a semi-strict feedback form; the resulting simulation findings substantiate our claims.