基于改进Pair-Copula方法的多风电场风速时空相关性建模  

Temporal-spatial correlation modeling of wind speeds in multiple wind farms based on improved Pair-Copula method

作　　者：陈凡[1] 王浩 刘海涛[1] 赵美莲[1] 张小莲[1] 高恒 Chen Fan;Wang Hao;Liu Haitao;Zhao Meilian;Zhang Xiaolian;Gao Heng(School of Electric Power Engineering,Nanjing Institute of Technology,Nanjing 211167,China;Xiangshui Power Supply Company,State Grid Jiangsu Electric Power Co.,Ltd.,Yancheng 224600,China)

机构地区：[1]南京工程学院电力工程学院,江苏南京211167 [2]国网江苏省电力有限公司响水县供电分公司,江苏盐城224600

出　　处：《可再生能源》2023年第1期60-66,共7页Renewable Energy Resources

基　　金：国家自然科学基金项目(52107098);南京工程学院科研基金项目(CKJB202001)。

摘　　要：传统Pair-Copula方法采用藤结构表示风电场风速之间的相依结构,能充分计及风电场风速之间的空间相关性,但缺乏风速时间尺度上相关性的考虑。针对此问题,文章提出了一种基于改进Pair-Copula的风速建模方法。该方法首先将藤结构中风电场风速的Copula联合分布函数样本点所包含的自相关信息转移至抽样模拟的随机数中,以此体现风电场风速数据在时间序列上的波动规律性。然后,采用分布更均匀的Sobol序列随机数代替普通随机数进行抽样,以提高风速概率特性模拟的准确性。最后,以位于美国东部的多个相邻风电场实测风速数据为例进行算例分析,算例结果验证了所提风电场风速建模方法的有效性。Vine structure is used to describe the dependence of wind speeds in multiple wind farms in the traditional pair-Copula method,which can fully consider the spatial correlation of wind speeds,but it ignores the temporal correlation of wind speeds.To solve this problem,a wind speed modeling method based on improved Pair-Copula method is proposed,which can transfer the autocorrelation information contained in the sample points of Copula joint distribution function of wind speeds in vine structure to the simulated random numbers,and thus can reflect the fluctuation characteristics of wind speeds in time series.Furthermore,a more uniform Sobol sequence is proposed to replace the routine random number sequence in sampling simulation,which improves the accuracy of probability characteristics of simulated wind speed.Finally,case studies have been carried out by using measured wind speeds of several adjacent wind farms located in eastern cities of American,and the results of case studies verified the effectiveness of the proposed modeling method.

关 键 词：PAIR-COPULA 风速 时空相关性 信息转移 Sobol序列 

分 类 号：TK81[动力工程及工程热物理—流体机械及工程]