F. Li, A COMPARISON OF GENETIC ALGORITHMS WITH CONVENTIONAL TECHNIQUES ON A SPECTRUM OF POWER ECONOMIC-DISPATCH PROBLEMS, Expert systems with applications, 15(2), 1998, pp. 133-142
Citations number
21
Categorie Soggetti
Computer Science Artificial Intelligence","Operatione Research & Management Science","Engineering, Eletrical & Electronic","Computer Science Artificial Intelligence","Operatione Research & Management Science
The mathematical formulation for a power economic dispatch problem can
be variously defined according to utilities' central focus. Conventio
nally, it is common practice to approach different dispatch problems v
ia different techniques. The time and design effort, thus induced in p
artly altering or entirely replacing the existing technique, is not de
sirable. This paper demonstrates the robustness of a search technique
based on genetic algorithms (GAs) against a number of conventional tec
hniques over a spectrum of power dispatch problems. The problems inves
tigated are in increasing order of complexity. Initially, GAs cannot d
o better than conventional techniques when the simple problem formulat
ion is encountered, e.g. in the case of static classic Economic Dispat
ch. However, when problems become progressively more complicated, GAs
gradually overtake conventional techniques which are limited mainly by
solution accuracy. The outcome of the study clearly shows the robustn
ess and suitability of GAs on the power dispatch problems, and verifie
s the fact that the more complex the problem is, the more benefit one
can obtain from a GA. (C) 1998 Elsevier Science Ltd. All rights reserv
ed.