An adaptive neighboring search using crossover-like mutation for multi modal function optimization

Abstract

We propose a new population-based Evolutionary Algorithm, which uses real-coded representation and normal distribution crossover-like mutation for generating next searching points. This Gaussian distribution is formed based on the position relationship between an individual and its neighbors, and is not carried with self-adapting parameters as inheritable traits. This algorithm causes emergence of clusters of individuals within the population, as the result of evolutions of each individuals without intent to cluster. Through searching independently, that emergent clusters introduce various solutions that include optimum at the same time, even if the problem has strong local minima. The proposed method robustly solves highly multimodal 30-dimensional Fletcher and Powell function by small population size.