A modified differential evolution for constrained optimization

This paper proposes a hybrid differential evolution method to solve constrained optimization problems. In the proposed method, we apply chaos theory to obtain self-adaptive parameter settings associated with DE. In order to handle constraint effectively, comparisons among feasible and infeasible solutions are made so as to provide a search direction towards the feasible region. Furthermore, a simple feasibility-based selection comparison technique is devised to guide the process toward the feasible region of the search space. The performance of the proposed method has been examined by its application to thirteen test cases, which were widely studied benchmark constrained optimization problems in literature. The results are very competitive with respect to three state-of-the-art evolutionary optimization algorithms and show that the proposed method is effective for handling constraints.