A Study on Model Reference Variable Structure Control with Adaptive Sliding Surface

Abstract

In this paper, for single-input single-output linear time invariant systems with unknown parameters, a new model reference variable structure control which has an adaptive sliding surface is proposed. This control is feasible using only input and output measurements. Under some usual assumptions in model reference adaptive control (MRAC) by applying the so-called e₁-modification1), the resulting closed loop system is shown to be globally stable in the presence of bounded input disturbances. Also disturbance attenuation is possible, to some extent, by changing the parameter of sliding surface. In addition, since the identification error model can be suitably changed without prefilter and the relative degree of it is smaller by 1 compared with ordinary MRACs, the tracking performance can be improved. Simulation results are presented to show these advantages.