On the Convergence of Imperialist Competitive Algorithm

Evolutionary algorithms have proved to be a powerful tool for solving complex optimization problems. Imperialist Competitive Algorithm (ICA) is a new evolutionary algorithm. Although ICA has been widely applied to solve many engineering problems, the convergence behavior of ICA is rarely discussed. This paper studies how ICA´s parameters affect its convergence behavior. Our results indicate that one can choose the parameters´ values intelligently to improve the exploration ability of ICA and still guarantee its convergence. Moreover, the results suggest the possibility of adaptive ICA that adjusts its parameters´ values dynamically to meet the need of diversity and convergence in the course of its execution.