FUZZY CONTROLLER - DESIGN, EVALUATION, PARALLEL AND HIERARCHICAL COMBINATION WITH A PID CONTROLLER

Fuzzy techniques still remain ill-accepted in the control community. A s a matter of fact, they rely on a new relation between the real world and the scientists. Whereas some theoretical studies are carried out on this subject, experiments on processes show what fuzzy techniques c an bring to the control theory. This paper deals with some implementat ions of control structures using fuzzy techniques. Our goal is to stud y the ways to include fuzzy techniques into existing control tools in order to improve the closed-loop performances. The experiments are car ried out on different processes with disturbances: a DC motor and indu strial furnaces. They allow to study the design and properties of stru ctures such as fuzzy control, commutation between fuzzy and PID contro llers and fuzzy supervision of PID controller. After having described two fuzzy controller implementations (fuzzy proportional-derivative an d fuzzy proportional-integral controllers), the comparison with a PID algorithm is a base for the design of the parallel PID-fuzzy controlle rs combination. Even though this strategy is interesting, another hier archical combination of PID and fuzzy tools is considered. The propose d fuzzy supervisor leads to promising results concerning the developme nt of combined control structures.