An adaptive NN controller with second order SMC-based NN weight update law for asymptotic tracking

Abstract

The asymptotic tracking control problem of a class of single-input single-output (SISO) uncertain nonlinear systems is addressed in this paper. A single-hidden layer neural network is used as a controller with a novel online weight training algorithm. The proposed NN weight update law mimics standard second order sliding mode control (2-SMC) approaches to ensure semi-global asymptotic convergence of the tracking error to the origin with continuous control effort. A simulation study verifies the effectiveness of the NN controller with 2-SMC-based online training. © 2009 Springer Berlin Heidelberg.