Feedforward control to track the output of a forced model

Abstract

In this paper a feedforward control technique is developed which enables a plant to perfectly track a model in the presence of a disturbance. It is assumed that both the model and the disturbance have time-varying inputs which are not known in advance. The technique used here develops dynamic compensators which have as inputs only a finite number of derivatives of the model and disturbance inputs. A procedure for using this technique in conjunction with optimal control is also given. Examples of how these teckniques can be used are presented.