Adaptive Neural Network Control For Nonlinear Uncertain Systems

In recent years, adaptive control is a nonlinear control methodology which is particularly useful for control of highly uncertain, nonlinear and complex systems. In the design and synthesis of nonlinear systems with uncertainty, the adaptive control based on backstepping approach has become the highlight in the control theory. Good control effects are acquired if NN are applied to uncertain systems. However this is a very general that a priori information about the practical systems may be given. Thus it is important to design controllers for uncertain system with various types of uncertainties.By combining adaptive neural controller with backstepping methodology, an adaptive neural network control scheme is presented based on Lyapunov's stability theory for a class of nonlinear systems in the strict-feedback form, together with unknown nonlinearities and unknown parameters. The priori information about the systems could be sufficiently exploited, much less neurons are employed for approximation, and the control performance of the closed-loop system is guaranteed by suitably choosing the design parameters. Furthermore, we give an extension of application for the control scheme. Simulation studies are conducted to verify the effectiveness of the proposed approach.The thesis consists of the following five chapters. The arts of the nonlinear adaptive control theory for uncertain system are stated in chapter one. In chapter two, some basic knowledge is presented. In chapter three, a direct adaptive neural network control is designed for a class of affine nonlinear systems in the strict-feedback form with unknown nonlinearities. In chapter four, an adaptive neural network control scheme is presented for a class of partially known nonlinear systems in the strict-feedback form. An extension of application for the control scheme is given. In the last chapter, some conclusions are drawn.