Evolutionary design of TSK fuzzy rule-based systems using (μ,λ)-evolution strategies

The main aim of this paper is to present an evolutionary process for designing TSK fuzzy rule-based systems based on the combination of an inductive algorithm that decides the number of rules forming the knowledge base, and a (μ,λ)-evolution strategy that determines their consequent parameters. Some aspects make this process different from others proposed till now: the use of an angular coding of the consequent parameters that allows us to search across the whole space of possible solutions, and the use of the available knowledge about the system under identification to generate the initial populations of the evolution strategies that allows us to speed up the search process, obtaining good solutions more quickly. The performance of the method proposed is shown by measuring the accuracy of the TSK fuzzy rule-based systems designed in the fuzzy modeling of two three-dimensional surfaces and comparing it with two Mamdani-type ones, generated by using inductive and evolutionary design processes