A novel diversification strategy for multi-objective evolutionary algorithms
Saul Zapotecas Martinez, Carlos A. Coello Coello · 2010
Diversity plays an important role in evolutionary multi-objective optimization. Because of this, the development of mechanisms which provide and maintain diversity to multi-objective evolutionary algorithms (MOEAs) have been studied since their inception. Fitness sharing and niching are probably the most popular density estimator used with non-elitist MOEAs. However, the main downside of these techniques is the need to define the niche radius, which is a critical parameter that is not trivial to set. In recent years, the use of external archives to store the non-dominated solutions found by elitist MOEAs has become a popular choice. This has triggered more effective and simple diversity maintenance techniques for MOEAs. In this paper, we introduce a new archiving strategy based on the Convex Hull of Individual Minima (CHIM), which is intended to maintain a well-distributed set of non-dominated solutions. Our proposed approach is compared with NSGA-II using standard test problems and performance measures taken from the specialized literature.