Adaptive discrete variable structure control

Abstract

The application of variable structure control (VSC) to various systems has increased rapidly in the last decade because of its known features such as robustness and implementation simplicity. Most VSC literature are in continuous time domain. However, expressing the VSC system in discrete time domain assists in suppressing many problems and obtaining new design methodology in the VSC using microprocessors. In this paper VSC is discretized and used with the ARMA model to estimate parameter plants which are used with the VSC reachability condition to determine the control-input signal. This method minimizes the stress on the system because the control-input doesn't depend upon the extreme boundaries of the plant parameters but depends instead upon their estimated values which minimizes the chattering in the control input signal. Also, another approach in which VSC is based on the ARMA model parameters, not plant parameters, has yielded significant results, which will have a profound effect upon the control of uncertain systems. Preliminary simulation results of this greatly simplified technique show the improved performance of the VSC system and the positive effects of varying the switching surface parameters on the system response.