Solving the economic dispatch problem with an integrated parallel genetic algorithm

Abstract

The application of an integrated parallel genetic algorithm (GA) incorporating simulated annealing (SA) and tabu search (TS) techniques to the economic dispatch (ED) problem is reported in this paper. The integrated genetic algorithm is implemented in both parallel and cluster structures. The parallel computing platform is based on a network of interconnected personal computers (PC) using TCP/IP socket communication facilities. Results from a case study of determining the optimal loading of 13 generators using a network of ten Pentium II-350 computers are presented. The proposed approach has the potential to be applied to other power engineering problem such as unit commitment and maintenance scheduling.