Iterative Design of a Mamdani Fuzzy Controller

Addressed to the problem of translating the control knowledge of a human expert operator into fuzzy control rules, this paper proposes an approach to automatically design a Mamdani fuzzy logic controller. The proposed approach is based on the use of a data set extracted from a process that has been manually controlled, and has the aim of learning a Mamdani logic controller with the capability to imitate the control action of an expert human operator. The proposed approach is an iterative method where fuzzy control rules can be added to the control knowledge base, as a function on the error between the target controller and the learned control. This iterative process stops when the learned controller reaches a behavior similar to the target controller or a maximum control structure complexity. Moreover, a real control setup composed by two coupled DC motors was used to test the performance of the proposed approach. The results show that the proposed approach has the capability of designing the Mamdani fuzzy controller in order to successfully controlling the real experiment.