Generalized theory for hybridization of evolutionary algorithms

Abstract

This study proposes to generalize the hybridization of evolutionary algorithm for solving large dimensional continuous global optimization problems. Inspired by various dual hybridizations being used, this paper proposes hybrid evolutionary algorithms based on crossing over the FFA, PSO, BAT, ACO and GA algorithms. The main idea of the proposed method is to integrate the aforementioned algorithms by following best solutions of other algorithm using roulette wheel approach. The aim of the proposed hybrid algorithm was to enable problem solving using two or more Evolutionary Algorithms as is, without modification, besides effectively exploring and exploiting of the problem search space. Simulations for a series of benchmark test functions justify that an adroit hybridization of various evolutionary algorithms could yield a robust and efficient means of solving wide range of global optimization problems than the standalone evolutionary algorithms.