An intelligent multi-controller structure implementation

In this paper we present a supervisory knowledge based control scheme which continuously monitors and makes a decision amongst several controllers running in parallel where each is automatically tuned online, using an identified plant model, to meet a number of user specifications. In its present form it includes a number of different control strategies such as: generalised minimum variance, pole placement control, general predictive control, and a proportional-integral-derivative (PID) controller, but the structure may be extended to contain as many controllers as deemed appropriate.