ペイオフの確率的特性を考慮した天候デリバティブの設計法

カテゴリ: 部門大会

論文No: 28

グループ名: 【B】平成27年電気学会電力・エネルギー部門大会

発行日: 2015/08/26

タイトル(英語): A Method for Designing Weather Derivatives with Probabilistic Characteristics of Payoffs

著者名: 藤田 創(明治大学),森 啓之(明治大学)

著者名(英語): Fujita Hajime(Meiji University),Mori Hiroyuki(Meiji University)

キーワード: 天候デリバティブ|最適化|契約モデル|DAクラスタリング|進化的計算|EPSO|Weather Derivative|Optimization|Contract Model||DA Clustering|Evolutionary CompotationEPSO

要約(日本語): In this paper, an efficient method is proposed for the design of weather derivatives with a hybrid intelligent system that consists of DA（Deterministic Annealing） clustering for global clustering and EPSO（Evolutionary Particle Swarm Optimization）of metaheuristics. The former is used to decompose data into some clusters while the latter works to evaluate globally optimal solutions. Weather derivatives play an important role to level the income for unexpected risk in case where weather conditions affect it. According to historical data, this paper deals with the contract of weather derivatives between an electric power and gas companies that may be expressed by nonlinear functions with parameters so that reasonable contract is created from a standpoint of optimization. The proposed method is successfully applied to real data

PDFファイルサイズ: 390 Kバイト