Economic Dispatch with Particle Swarm Optimization for Large Scale System with Non-smooth Cost Functions Combine with Genetic Algorithm

Abstract

ABSTRACT This paper presents a novel stochastic optimization approach to determine a feasible optimal solution for the economic dispatch (ED) problems, considering various generator constraints. Many practical constraints of generators, such as ramp rate limits and prohibited operating zones are investigated. In order to improve the performance of the particle swarm optimization (PSO) algorithm. These constraints alter the ED problem to a non-smooth minimization problem with constraints, an innovative approach based on PSO is chosen to solve the load-flow problem by combining the genetic algorithm (GA), i.e. some of the answers were found by the PSO algorithm itself; after that two possible procedures are studied: one is utilizing the PSO algorithm to find the optimum answer among the primary guesses and the other is using a GA implementation by means of the arithmetic crossover operators. The last step is choosing the best answer among the obtained results (GA, PSO). To show its efficiency and effectiveness, the proposed algorithm (GPSO) is applied to some types of ED problems containing non-smooth cost functions of 13 and 40 generating units systems (large scale systems). The experimental results show that the GPSO approach is comparatively capable to obtain higher quality solution.