On the centrality of disturbance rejection in automatic control.

In this paper, it is shown that the problem of automatic control is, in essence, that of disturbance rejection, with the notion of disturbance generalized to symbolize the uncertainties, both internal and external to the plant. A novel, unifying concept of disturbance rejector is proposed to compliment the traditional notion of controller. The new controller-rejector pair is shown to be a powerful organizing principle in the realm of automatic control, leading to a Copernican moment where the model-centric design philosophy is replaced by the one that is control-centric in the following sense: the controller is designed for a canonical model and is fixed; the difference between the plant and the canonical model is deemed as disturbance and rejected.