Model-following VSC using an input-output approach

Abstract

The model-following problem for uncertain linear SISO systems in input/output form has been discussed. The approach presented is characterized by the use of a discontinuous control law within a control structure inherited from classical adaptive control schemes and simplified, since its complexity turns out to be independent of the relative degree of the plant to be controlled.

The structure of the linear scheme corresponding to the ideal case of perfect knowledge of the plant parameters was first described. The main feature of the design procedure underlying the proposed scheme consists of the fact that a pole assignment objective is pursued by performing explicit model tracking and synthesizing a suitable feedforward compensator.

In the case of an uncertain plant the two stages into which the design process can be divided, were considered in sequence. The tracking problem is easily dealt with by using a discontinuous parameter adjustment mechanism based on the knowledge of the uncertainty bounds of the plant parameters. The pole assignment problem is solved by applying an identification procedure and exploiting the link between the parameters of the feedforward compensator and those of the feedback part of the controller.

The advantage of the described approach with respect to other proposals in the literature mainly relies on its simplicity and on the fact that it is based on a reduced assumption set about plant uncertainty.