Abstract
Fuzzy reasoning models are model-free estimators of control systems; this makes them very powerful tools in control applications. Their performance depends on several factors such as the completeness of the rule base, the subjective definitions of the fuzzy subsets parameters or fuzzy partitioning, the fuzzy implication operator, and the defuzzification method.
In most cases, these factors are decided upon subjective experience of the human expert.
In this paper, we attempt to improve the fuzzy systems performance by means of genetic-based learning mechanisms.
We propose two methodologies. These are two-step optimization methods. In both methodologies, a Genetic Algorithm (GA) loop is used for parameter optimization and selection, and competitive clustering to generate clusters.
One methodology uses a supervised training algorithm for fine tuning the fuzzy structure (GA-ST), while the other uses an inner GA-loop to derive optimal Fuzzy Reasoning Model (FRM) parameters (GA(2)).