基于Vine-Copula和拟蒙特卡罗法的可再生能源电力现货市场风险度量模型

Renewable energy electricity market risk estimating model based on Vine-Copula and Quasi-Monte Carlo method

可再生能源参与电力现货市场政策制定前,需度量可再生能源的市场风险。在可再生能源未深度参与电力现货背景下,直接采用传统方法对历史数据进行风险度量不可行。文章结合可再生能源有效出力和在险模型(Value at Risk,VaR)设计了可再生能源风险度量指标,通过风险场景匹配市场和机组出力数据。针对目前国内电力现货市场数据量少的问题,结合市场主体在电力现货市场中所能获得的数据进行市场因子筛选,通过Vine-Copula函数考虑多个市场因子间的相关性;改进了传统蒙特卡罗法收敛慢的缺陷,采用拟蒙特卡罗模拟法生成市场因子数据,根据拟合的映射关系生成电价水平数据。最后,文章基于南方(以广东起步)电力现货市场结算试运行的数据和海上风电与光伏的仿真出力,对所提出的模型进行算例分析,结果显示拟蒙特卡罗法收敛性更高,能满足模型中所有风险场景高频计算风险的需求。

The reform of electricity spot markets and dual carbon policies permits renewables participating in spot markets to cul-tivate price discovery and the replacement of conventional energy sources.Consequently,prior to formulating policies surrounding renewables'involvement in spot markets,the market risks of renewables must be assessed.Under current circumstances where renewables have not been deeply incorporated into spot markets,and employing historical data directly for risk measurement is no longer suitable,this paper amalgamates the effective output of renewables with VaR models to conceive renewables risk metrics that correspond with market and unit output data through risk scenar-ios.To address the presently small amount of domestic spot market data accessible to market participants,market factors are screened by utilizing the data entities can obtain,characterizing correlations using Vine-Copula functions.Instead of traditional Monte Carlo simulations with slow convergence,quasi-Monte Carlo simulation methods are em-ployed to generate market factor data,from which electricity price levels are derived according to fitted mapping rela-tionships.Ultimately,based on pilot program data from Guangdong spot market and simulation outputs of offshore wind and photovoltaic power,the analysis indicates quasi-Monte Carlo methods exhibit higher convergence and can satisfy the high-frequency risk calculation requirements across all risk scenarios.