This paper suggests a framework based on participatory learning and evolutionary computation to develop fuzzy models. Participatory evolutionary learning is an algorithm in which the population influences the fitness of individuals during evolution. The framework selects individuals for reproduction using the compatibility between the best and individuals randomly selected from the old and current population. Recombination uses selective transfer to exchange information between individuals, and mutation proceeds similarly as mutation in differential evolution. Selective transfer and selection use compatibility information between individuals, and mutation uses an arousal index in addition to compatibility. Combination of participatory learning with selection, selective transfer and mutation offers a new class of evolutionary algorithms that are particularly useful in fuzzy modeling. The paper emphasizes model development for an application concerning electric system maintenance. The application uses actual data available in the literature, and helps to compare the performance of participatory evolutionary learning with an alternative participatory genetic learning, and with a state of the art genetic fuzzy system tool. Modeling performance is evaluated using the mean squared error and number of rules to measure model accuracy and complexity, respectively. The results show that participatory evolutionary learning is competitive because it produces fuzzy models with similar complexity, better accuracy, and low processing time.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.