Adaptive variable structure tracking control using neural network design

This paper presents an adaptive tracking control approach to linear SISO systems, which can solve the traditional model reference adaptive control (MRAC) problems. A new error model is developed for design of an adaptive variable structure controller using only input-output measurements. In this approach, a neural network universal approximator is included to furnish an on-line estimate of a function of the state and some signals relevant to the desired trajectory. It is shown via Lyapunov stability theory that the asymptotic tracking accuracy of the closed-loop system can be arbitrarily improved by decreasing a positive design parameter r, whose inverse characterizes the bandwidth of a so-called averaging filter.