基于Copula函数的多维时序风速相依模型及其在可靠性评估中的应用

A Copula Function-Based Dependent Model for Multivariate Wind Speed Time Series and Its Application in Reliability Assessment

风速模拟是对含风电的电力系统进行可靠性评估的先决条件。基于Copula函数和马尔科夫过程理论,提出了一种多维时序风速相依模型及该模型的两阶段蒙特卡洛模拟方法。该模型将多维相依关系分解为单维风速时间序列的短期相依关系和多个风速序列之间的同期相依关系。利用1阶连续状态马尔科夫链描述风速时间序列的变化规律;利用Copula函数捕捉风速序列中存在的短期相依关系和同期相依关系。通过实际风速数据对模型进行了验证,结果表明提出的模型能够较好保持风速序列的概率分布、自相关和互相关特性。将该模型应用到IEEE-RTS可靠性测试系统中,结果表明上述模型能够直接用于风电场可靠性效益评估。

Wind speed simulation is a prerequisite to reliability assessment of power grid containing wind farms.Based on Copula function and Markov process,a dependent model for multivariate wind speed time series（WSTS） is proposed and a two-stage Monte Carlo simulation method for the proposed model is given.In the proposed model,the multivariate dependence is decomposed into temporal dependence of univariate WSTS and contemporaneous dependence among multi WSTS.The first-order continuous state Markov process is utilized to describe the variation of WSTS and the Copula function is utilized to capture temporal dependence and contemporaneous dependence existing in WSTS.The proposed model is verified by actual wind speed data,and verification results show that in the proposed model the probability distribution as well as autocorrelation and cross-correlation performances of WSTS can be well maintained.Applying the proposed model to IEEE-RTS system,test results show that the proposed model can be directly applied in reliability assessment of power grid containing wind farms.