Authors: Khoa Duc Tran
Addresses: KDT Consulting Group, 9541 Mansor Avenue, Garden Grove, California, USA
Abstract: Multi-Objective Evolutionary Algorithms (MOEAs) are not easy to use because they require parameter tunings to achieve good solutions and performance for an arbitrary complex problem. This paper introduces a MOEA with adaptive population size, self-adaptive crossover and self-adaptive mutation for automating the process of adjusting parameter values to make the MOEA simple to use. The new MOEA is built on the Non-dominated Sorting Genetic Algorithm-II (NSGA-II) and named as Adaptable NSGA-II (ANSGA-II). Simulation results on 13 multi-objective problems demonstrate that the ANSGA-II out-performs the NSGA-II in terms of finding diverse non-dominated solutions and converging to the true Pareto-optimal front.
Keywords: genetic algorithms; GAs; multi-objective evolutionary algorithms; MOEA; multi-objective optimisation; parameter control techniques.
International Journal of Intelligent Systems Technologies and Applications, 2009 Vol.7 No.4, pp.347 - 369
Available online: 03 Sep 2009 *Full-text access for editors Access for subscribers Purchase this article Comment on this article