Improved Imperialist Competitive Algorithm for Constrained Optimization

This paper introduces an improved evolutionary algorithm based on the imperialist competitive algorithm. The original approach in the imperialist competitive algorithm has difficulty in implement practically with the increase of the dimension of the search spaces, as the ambiguous definition of the ¿random angle¿ in the process of optimization. Compare to the original algorithm, the proposed approach based on the concept of small probability perturbation has more simplicity to be implemented, especially in solving high-dimensional optimization problems. Furthermore, the present algorithm has been extended to constrained optimization problem, using a classical penalty technique to handle constraints. Several numerical optimization examples are tested by applying the proposed algorithm, and the results show its applicability and flexibility in dealing with different types of optimization problems.