where r1, r2are random numbers uniformly distributed in [0,1]. ωis the inertia weight, C1and C2are acceleration constants, and d isthe dth dimension of the search domain.Although PSO has fast convergence behavior, there is some defi-ciency in PSO performance. This is due to that all particles learnfrom gbest. If gbest falls in local optima the whole swarm may followit and this would lead to premature convergence.Various attempts have been made to overcome the shortcomingof PSO. One of these attempts is hybridizing PSO with other searchtechniques [10–17] to benefit from advantages of each algorithm.The hybridization between DE [18] and PSO looks a promising opti-mizer. This hybridization tries to benefit from good global searchcapability of DE and high speed convergence of PSO.DE is a simple evolutionary algorithm for global optimizationproposed by Price and Storn [18]. The DE-variants perturb the cur-rent population members with the scaled differences of randomlyselected and distinct population members. Therefore, no separate