A General PSO Modification for Electrical Applications

Particle swarm optimization technique (PSO) is a promising evolutionary technique. Setting initial values of that technique depend on every domain of application. These values are chosen randomly in classic PSO technique which increases the computation time. This paper presents a new approach to modify PSO based on a prediction technique to determine the PSO initial values and number of particles for each optimization problem. The proposed modification makes the PSO technique faster and more applicable for electrical online applications. The new approach is used to solve the protective devices coordination problem by choosing the optimum initial values for the coordination domain. A comparison between the classic PSO and the proposed modified one shows that the new approach gives fast results and shortens the computation time. Keywords-Evolutionary Techniques, Particle Swarm Optimization (PSO), Coordination, DG.