A hybrid Differential Evolution with double populations for constrained optimization

How to balance the objective and constraints is always the key point of solving constrained optimization problems. This paper proposes a hybrid differential evolution with double populations (HDEDP) to handle it. HDEDP uses a two-population mechanism to decouple constraints from objective function: one population evolves by differential evolution only according to either objective function or constraint, while the other stores feasible solutions which are used to repair some infeasible solutions in the former population. Thus, this technique allows objective function and constraints to be treated separately with little costs involved in the maintenance of the double population. In addition, to enhance the exploitation ability, simplex method (SM) is applied as a local search method to the best feasible solution of the first population. Simulation results based on three well-known engineering design problems as well as comparisons with some existed methods demonstrate the effectiveness, efficiency and robustness of the proposed method.