Evaluation of a Multi-Variable Self-Learning Fuzzy Logic Controller
In spite of the usefulness of fuzzy control, its main drawback comes from lack of a systematic control design methodology. The most challenging aspect of the design of a fuzzy logic controller is the elicitation of the control rules for its rule base. In this paper, a scheme capable of elicitation of acceptable rules for multivariable fuzzy logic controllers is derived by extending an algorithm that enables a single-input-single-output fuzzy logic controller to self-learn its rule-base. The performance of the proposed self-learning procedure is investigated and evaluated by means of simulation studies of a hypothetical plant. The results obtained indicate that the approach could be effective in the control of nonlinear multivariable industrial processes.