Solving practical economic dispatch using hybrid GA–DE–PS method

Abstract

This paper presents an efficient flexible hybrid method based two global optimization methods such as; genetic algorithm, differential algorithm dynamically coordinated with a local search method known as pattern search to enhance the solution of the economic dispatch problem considering valve point effects. The main idea introduced in this proposed hybrid method is that the best solutions found by GA and DE at different stages are sent to PS to exploit new region around this solution, alternatively the new solution achieved by PS is also communicated to GA and DE and considered as the new initial search space, this interactive mechanism search between global and local search is to balance the exploitation and exploration capability which allows individuals to react more by changing experiences. The robustness of the proposed approach is tested and validated on the Algerian 114-bus electrical network test by solving the economic dispatch considering quadratic cost function and fixed power losses, to the 13 and 40 generating units considering valve-point effects. Comparison results with the standard global optimization methods such as GA, DE and to many other recent techniques showed the superiority of the proposed hybrid technique and confirmed its potential for solving practical economic dispatch.