加权CVaR下的发电商多时段投标组合模型  

Analysis of Multi-period Combined Bidding of Power Suppliers Based on Weighted CVaR

作　　者：张兴平[1] 陈玲[2] 武润莲[1] 

机构地区：[1]华北电力大学工商管理学院,北京市昌平区102206 [2]泉州电业局,福建省泉州市362000

出　　处：《中国电机工程学报》2008年第16期79-83,共5页Proceedings of the CSEE

基　　金：国家自然科学基金项目(70671042);教育部人文社会科学研究项目(06JC790014)~~

摘　　要：在多市场条件下,为了实现收益最大和风险最小,发电商需要在不同市场合理分配竞价电量。发电商竞价决策是多阶段决策,因而决策过程常呈现多期风险,即动态风险。同时各类电力市场收益率并非固定不变,而是具有随机变化的特性。条件风险价值(conditionalvalueatrisk,CVaR)可以衡量决策过程的风险和收益,而单时段CVaR只考虑单一时段的静态风险和固定收益率,因此采用多时段CVaR模型度量发电商的动态风险和随机变化的收益率,并建立发电商加权多时段CVaR组合市场投标策略优化模型。应用该模型,发电商可以在不同时段将竞价电量在多市场中进行合理分配,以实现风险最小和收益最大。算例分析结果表明该方法的有效性,从而为发电商的投标决策和风险度量提供新的思路。In the multi-market, power suppliers should allocate the bidding electricity to each market reasonably in order to minimize the risk and maximize the profits. Decision-making is a multi-period process in bidding, so power suppliers face the multi-period risks （dynamic risks）; the revenue rates of different markets have characteristics of stochastic changing. The model of conditional value at risk （CVaR） can measure the risks and revenues, but single period CVaR only reflects the static risk and the fixed revenue rate of single period. So the multi-period bidding strategy model based on the CVaR is applied in bidding, which can rationally reflect the dynamic risks and the stochastic changing characteristics of revenue rates. Applying the model, power suppliers allocate the bidding electricity volume in different periods and different markets rationally to minimize the risk and maximize the expect profits. The calculation results show the validity of the multi-period CVaR model, and proviaes a new way for bidding strategy and risk valuation.

关 键 词：电力市场 投标组合 风险计量 多时段条件风险价值 

分 类 号：TM73[电气工程—电力系统及自动化] F123[经济管理—世界经济]