Development of Bidding Strategies in a Competitive Market Using Artificial Intelligence

• Main Authors: Shin-Wen Tsai, 蔡興文
• Other Authors: Ying-Yi Hong
• Format: Others
• Language: zh-TW
• Published: 1999
• Online Access: http://ndltd.ncl.edu.tw/handle/02227556435308579785

碩士 === 中原大學 === 電機工程學系 === 87 === The structure of the power industry has being changed from monopoly to competition. This thesis analyzes the recent development of deregulated electric power markets and figures out the new organizations, Independent System Operator (ISO) and Power Exchange (PX). After deregulation, purchase/sales of the electricity becomes a commercial transaction. Each genco offers a bid price and the ISO takes the all bidding prices with network constraints into account to determine the generation scheduling. In this thesis, a bidding strategy is developed for a genco in a competitive market. A bidding price and the status of win/lose are obtained from the proposed bidding strategy. The information of the nodal prices is assumed to be available in the competitive market for the participants in the paper. An artificial neural network (ANN) based on the nodal prices using the back-propagation algorithm is developed for bidding. Fuzzy-c-Mean (FCM) algorithm is employed to cluster the load levels and the gecos for reducing the training time for the ANN. The impact of different network combinations on the training time is investigated. The IEEE 30-bus system is used to be an example for generating scheduling data. The success rates for different ANN with different load-level/genco combinations are explored. A best combination with load levels/generators is suggested for a genco to the proposed bidding strategy.