Mechanical Component Design for Multiple Objectives Using Elitist Non-dominated Sorting GA

In this paper, we apply an elitist multi-objective genetic algorithm for solving mechanical component design problems with multiple objectives. Although there exists a number of classical techniques, evolutionary algorithms (EAs) have an edge over the classical methods in that they can find multiple Pareto-optimal solutions in one single simulation run. The proposed algorithm (we call NSGA-II) is a much improved version of the originally proposed nondominated sorting GA (NSGA) in that it is computationally faster, uses an elitist strategy, and it does not require fixing any niching parameter. On four mechanical component design problems borrowed from the literature, we show that the NSGA-II can find a much wider spread of solutions than classical methods and the NSGA. The results are encouraging and suggests immediate application of the proposed method to other more complex engineering design problems.