Engineering case studies using parameterless penalty non-dominated ranked genetic algorithm

Abstract

The new elitist multi-objective genetic algorithm PPNRGA have been used for solving engineering design problems with multiple objectives. Although there exists a number of classical techniques, evolutionary algorithms (EAs) have an edge over the classical methods where they can find multiple Pareto optimal solutions in one single simulation run. The new proposed algorithm is a parameterless penalty non-dominated ranking GA (PP-NRGA), uses a fast non-dominated sorting procedure, an elitist-preserving approach, a two tier ranked based roulette wheel selection operator, and it does not require fixing any niching parameter. PP-NRGA tested on two engineering design problems borrowed from the literature, where the PP-NRGA can find a much wider spread of solutions than NSGA-II other evolutionary algorithm. The results are encouraging and suggests immediate application of the proposed method to other more complex engineering design problems. © 2009 IEEE.