Application of Metaheuristics to Energy Supply Systems

カテゴリ: 部門大会

論文No: OS7-12

グループ名: 【C】平成20年電気学会電子・情報・システム部門大会講演論文集

発行日: 2008/08/20

タイトル(英語): Application of Metaheuristics to Energy Supply Systems

著者名: 北村 聖一(三菱電機),森 一之(三菱電機),泉井 良夫(三菱電機)

著者名(英語): Shoichi Kitamura(Mitsubishi Electric Corporation),Kazuyuki Mori(Mitsubishi Electric Corporation),Yoshio Izui(Mitsubishi Electric Corporation)

キーワード: metaheuristics|optimization|energy supply system|PSO|multiple solutions

要約(日本語): In many engineering applications, it is necessary to find multiple solutions to an optimization problem with complex multi dimensional objective function. Energy supply systems for example have better controlling systems when they can choose, among multiple solutions, the most attractive solution in cost but also a solution near to the actual regime. Pair PSO (Particle Swarm Optimization) is a successful algorithm for detecting multiple solutions of two dimensional problems. Algorithms based on particle swarm optimization for detecting multiple solutions in multi dimensional problems are developed in this paper. Multi PSO is a repetitive algorithm and shows good results in an acceptable time limit. The parallel searching Niche PSO algorithm can find more solution in the same time, but the precision of solutions is slightly inferior. Multi PSO and Niche PSO are simple algorithms which are able to solve multidimensional problems and detect multiple solutions. The goal in this approach is to detect solutions with high fitness values, which are determined by the objective function. The aim is not to detect local optima with low fitness value.

PDFファイルサイズ: 3,201 Kバイト