Neural network-based adaptive dynamic surface control of uncertain nonlinear pure-feedback systems

In this paper, by incorporating the dynamic surface control technique into a neural network-based adaptive control design framework, we have developed a backstepping-based control design for a class of nonlinear systems in pure-feedback form with arbitrary uncertainty. The circular design problem which may exist in pure-feedback systems is overcome. In addition, our development is able to eliminate the problem of 'explosion of complexity' inherent in the existing backstepping-based methods. A stability analysis is given, which shows that our control law can guarantee the semi-global uniformly ultimate boundedness of the solution of the closed-loop system, and makes the tracking error arbitrarily small. Moreover, the proposed control design scheme can also be directly applied to the strict-feedback nonlinear systems with arbitrary uncertainty. Copyright (C) 2010 John Wiley & Sons, Ltd.