Particle Swarm Optimization based Algorithm for Economic Load Dispatch

Particle Swarm Optimization (PSO) has emerged as a useful tool for engineering optimization. However, most of the PSO algorithms are aimed at unconstrained problems. For the constrained problems, the approach introduced is just the traditional combination of primitive PSO and the penalty function. It was investigated that the simple penalty function strategy cannot be integrated well with PSO algorithms because it does not utilize the historical memory information, which is an essential of PSO. In view of this problem, a new constraints handling strategy suitable for the optimization mechanism of PSO is proposed. In addition, local search procedure is hybridized with PSO to intensify its search ability in local regions, and a hybrid PSO (HPSO) algorithm for constrained optimization is put forward. Economic load dispatch problem is used to demonstrate its efficiency. Experimental results show that the proposed HPSO algorithm is both effective and efficient.