自由市場下發電業競標策略之研究

The GenCo's Bidding Strategy in a Deregulated Environment

蔡明堂(Ming-Tong Tsay) ； 郭宏傑(Hong-Jey Gow) ； 楊進德(Chin-Der Yang)

《永達學報》 8卷2期 (2007/12) Pp. 56-67

摘要

本文提出改良免疫演算法之搜尋方式，探討輸電線路壅塞環境下，發電業者的最佳競標策略，以謀求自身最大的利潤。本文首先利用參與業者競標標單的撮合，求得市場的結清價格及結清容量，並求得各參與業者的得標發電量。其次以電力追蹤的數學模式求得各發電機對線路的貢獻量，當線路發生壅塞時，將以比例調度法電量調度規則，實施發電增減量的策略，直至線路壅塞情形改善為止；在發電業者的競標策略上，結合最小誤差法及免減量意願因子，推導發電業者競標策略的模式，並以本文提出的改良免疫演算法求得各發電業因線路壅塞對競標策略的影響，以及計算發電業者的損失或獲利。本文亦以基因演算法及免疫演算法模擬競標策略的模式，以收斂時間及收斂解比較本文提出方法之優缺點及可行性。透過本文的執行，將提供市場參與者當面臨線路壅塞時，評估自身運轉安全及成本概況，執行精確的競價策略以獲得最大的利潤。

關鍵字

電力市場 ； 競標策略 ； 線路壅塞 ； 免疫演算法

並列摘要

This paper presents an Ehanced Immune Algorithm(EIA) to solve the optimal strategy of participants wishing to maximize its profit in congestion environments. In this paper, the bids accumulated method is used to find the market clear price and market clear capacity, and the generating output of participants. The contributed flow of each generator in a line will be calculated by using power flow tracing. When the power system has the congestion after the bids, the proportional dispatch method will be used to regulate the power output to meet the security constraints. For the bidding strategy of participants, the minimal square-error deviation and willingness factors are combined to form the model of bidding strategy which will be solved by the proposed algorithm. A bidding strategy will be also used in the bidding process to obtain the generators' profit/loss. The profit's deviations of congestion's influence for all participants are analyzed in detail. Numerical analysis will clarify congestion's influence on price and bidding strategy. Comparing the genetic algorithm, immune algorithm, and the proposed AI method, the converging solution and converging time will be obtained after bidding process. Simulation results will provide the participants to make possible an anticipated and optimal control strategy to obtain the maximal profit under participants' cost and security operation.

並列關鍵字

Power Market ； Bidding Strategy ； Congestion ； Immune Algorithm