Otimização por enxame de partículas em problemas de despacho econômico de carga

Abstract

This work aims to present the application of the Particle Swarm Optimization algorithm (PSO), using a stopping criterion, in the solution of the Economic Load Dispatch (ED) problem, considering some of the main constraints presented on a real ED model. It considers case studies consisting of thermoelectric systems with three and fifteen generating units, taking into account losses in the transmission lines, power capacity limits and prohibited zones of operation. The proposed algorithm is tested on case studies proposed in the literature, with varying degrees of difficulty. The solution obtained in this way is then compared with results in the literature, which uses PSO without stopping criterion, and with the exact solution of the problem obtained via the Lagrange Multipliers Method (LMM). The results obtained through computer simulations showed that the PSO with stopping criterion finds the exact optimal solution of the problem, or, in more complex problems, approaches substantially the solution given by LMM. In addition, when compared to PSO without stopping criteria, the proposed method converges giving a lower total generation cost, and lower losses in the transmission lines, converging faster than the method used in the literature for the case with three generation units.