Dynamic Fuzzy Logic Control of Genetic Algorithm Probabilities

Genetic algorithms are commonly used to solve combinatorial optimization problems. The implementation evolves using genetic operators (crossover, mutation, selection, etc.). Anyway, genetic algorithms like some other methods have parameters (population size, probabilities of crossover and mutation) which need to be tune or chosen. In this paper, our project is based on an existing hybrid genetic algorithm working on the multiprocessor scheduling problem. We propose a hybrid FuzzyGenetic Algorithm (FLGA) approach to solve the multiprocessor scheduling problem. The algorithm consists in adding a fuzzy logic controller to control and tune dynamically different parameters (probabilities of crossover and mutation), in an attempt to improve the algorithm performance. For this purpose, we will design a fuzzy logic controller based on fuzzy rules to control the probabilities of crossover and mutation. Compared with the Standard Genetic Algorithm (SGA), the results clearly demonstrate that the FLGA method performs significantly better. Yi Feng Degree Project May 2008 E3601D _______________________________________________________________________________ Högskolan Dalarna Tel: +46-23-778800 Röda Vägen 3, 781 88 Fax: +46-23-778050 Borlänge, Sweden Http://www.du.se Acknowledgement Before accessing this thesis, I would like to express my gratitude to my supervisor, Dr. Pascal Rebreyend, for his constant and invaluable guidance, encouragement throughout all phases of my research work. I also like to express my sincere thanks to Prof. Mark Dougherty, Dr. Hasan Fleyeh and Dr. Yella Siril. I learn a lot from your teaching. You are the best teacher for me in Sweden. Thanks to ShengTong Zhong, Yang Shen, and Ping Zhao. You were, are, and always will be my best friends. Special thanks to you, you give me the most help on my very hard time. Yi Feng Degree Project May 2008 E3601D _______________________________________________________________________________ Högskolan Dalarna Tel: +46-23-778800 Röda Vägen 3, 781 88 Fax: +46-23-778050 Borlänge, Sweden Http://www.du.se Dedicated...