Fuzzy Logic Controller Based on Genetic Algorithm for Electric Drive of Crane Trolley

Abstract

The paper describes a methodology based on the genetic algorithm that permits to automate the process for obtaining optimum rule bases while applying a controller based on a fuzzy logic. Using a specific structure of  a chromosome, a special mutation operation and an adequate fitness function  the proposed methodology makes it possible to select a fuzzy rule base, to minimize a number of rules, rationally to place input sets of fuzzy functions and corresponding arrangement of output singletons in the form of single-element sets. The paper presents an example when the methodology allows to reduce a number of rules by two fold and, consequently, calculative control procedure without any deterioration in quality control.