부산대학교 도서관

Since the 1960's a great deal of attention has been paid to the problem of reactive power compensation in electrical networks. This problem consists in providing the appropriate location and sizing of compensation devices, aimed at ensuring satisfactory voltage profile and reduce power losses, while minimizing the compensation cost. In recent years the methodologies to solve this problem shifted from mathematical programming algorithms to metaheuristics and, in particular, to Evolutionary Algorithms (EAs). EAs can provide good quality solutions without requiring mathematical model simplifications. Moreover, EAs are well suited to multi-objective problems because they work with a population of solutions. This paper makes an overview of the application of multi-objective EAs to the reactive power planning problem.