Fuzzy control and an evaluation of the self-organizing fuzzy controller

Fuzzy control is a rule based type of control that aims to imitate the human's ability to express a control policy using linguistic rules, and to reason using those rules to control a system. Fuzzy control is nonlinear and not dependent on a precise mathematical description of the plant, and is therefore more easily applied to systems such as industrial processes that are hard to model. An overview is given of the fuzzy controller, along with descriptions of applications and theoretical approaches to designing and analyzing the controller.

The self-organizing controller is able to generate or modify its rules in real time based on the system performance. It was tested to determine how well it was able to learn a quality control policy. The self-organizing controller was found to exhibit poor steady state performance, and to be equally likely to learn poor control as to learn good control. It was not found to eliminate the need for careful tuning of the controller parameters and gains.