Fuzzy logic versus niched Pareto multiobjective genetic algorithm optimization
Brian Reardon · Modelling and Simulation in Materials Science and Engineering · 1998
A new multiobjective selection procedure for a genetic algorithm (GA) based on the paradigms of fuzzy logic is introduced, discussed and compared to the niched Pareto selection procedure. In the two example problems presented here (Schaffer's F2 problem and a simplified Born-Mayer potential) the fuzzy logic procedure optimized the parameters of functions in a manner of comparable efficiency to that of the niched Pareto approach. The two main advantages of the fuzzy logic approach over the niched Pareto approach are that the experimental error or `uncertainty' in the objective values can be accounted for and, unlike the niched Pareto approach, the efficiency of the fuzzy logic GA is shown to be independent of the number of objectives.