基于改进云分段模型的光伏功率缺失数据补齐研究  

Research on photovoltaic power missing data completion based on improved cloud segmentation model

作　　者：张弘鹏[1] 刘家庆[1] 段志伟 徐志英 方渊[2] Zhang Hongpeng;Liu Jiaqing;Duan Zhiwei;Xu Zhiying;Fang Yuan(Northeast Branch of State Grid Corporation,Shenyang 110180,China;Guodian Nanrui Technology Co.,Ltd.,Nanjing 211106,China)

机构地区：[1]国家电网公司东北分部,辽宁沈阳110180 [2]国电南瑞科技股份有限公司,江苏南京211106

出　　处：《可再生能源》2020年第12期1590-1596,共7页Renewable Energy Resources

基　　金：国家自然科学基金(51307051);国家电网公司东北分部科技项目(52992618009Q)。

摘　　要：光伏功率缺失数据阻碍了对光伏功率的相关研究,因此,须要对光伏功率缺失数据进行有效补齐。在分析各种补齐方法的基础上,文章对原始光伏功率数据进行了描述和预处理,依据光伏功率模型对光伏功率数据进行分段,利用正向云算法,建立并实现了传统云分段模型。考虑到传统云分段模型中正态随机熵的不足,文章利用联合经验函数拟合出各分段太阳辐照度与光伏功率的联合分布,得到改进的随机熵并建立改进的云分段模型,参照云分段模型结果和不同光伏功率缺失数据的波动特性,构建了光伏功率缺失数据的补齐模型。通过不同方法对不同种类和比例的缺失数据进行补齐,分析结果表明,文章所提出的光伏功率缺失数据补齐方法的效果优于滑动平均法和传统云分段模型。Missing data of PV(photovoltaic)powerhinders relevant researches on PV power.Therefore,it is necessary to effectively supplement missing data of PV power.Based on the analysis of various complementation methods,the paper described and preprocessed the original PV data,segmented the data according to the PV power model,established and realized the traditional cloud segmentation model by using the forward cloud algorithm.Considering the traditional cloud segmentation model is the shortage of the random entropy,as a result,the article used of combined experience in function fitting the joint distribution of each block irradiance and power improved stochastic entropy and improved cloud segmentation model was set up,with reference to the cloud segmentation model results and the wave characteristics of different missing data,built the PV power model of filling missing data.Different types and proportions of missing data were complemented by different methods.The analysis results showed that the complementing effect of the proposed method for complementing missing photovoltaic power data was better than that of the moving average method and traditional cloud segmentation model.

关 键 词：光伏功率 缺失数据补齐 云分段模型 联合经验函数 

分 类 号：TK51[动力工程及工程热物理—热能工程] TM614[电气工程—电力系统及自动化]