文章摘要
杨 茂,黄鑫.基于光照过程特征分析的光伏功率异常数据识别算法[J].电力系统自动化. DOI: 10.7500/AEPS20180626003.
YANG Mao,HUANG Xin.Abnormal Data Identification Algorithm for Photovoltaic Power Based on Characteristics Analysis of Illumination Process[J].Automation of Electric Power Systems. DOI: 10.7500/AEPS20180626003.
基于光照过程特征分析的光伏功率异常数据识别算法
Abnormal Data Identification Algorithm for Photovoltaic Power Based on Characteristics Analysis of Illumination Process
DOI:10.7500/AEPS20180626003
关键词: 光伏功率异常数据  光照特征  Copula函数  区间最窄原则
KeyWords: Photovoltaic power abnormal data  illumination characteristics  Copula function  the narrowest interval
上网日期:2019-01-31
基金项目:国家重点研发计划项目’促进可再生能源消纳的风电/光伏发电功率预测技术及应用’(2018YFB0904200)
作者单位E-mail
杨 茂 东北电力大学电气工程学院 yangmao820@163.com 
黄鑫 东北电力大学电气工程学院 15044674026@163.com 
摘要:
      高质量的光伏功率数据是进行光伏研究的基础,而从光伏电站采集到的数据含有较大比例的异常数据,因此需对光伏功率异常数据进行识别。将不同光照特性的光伏功率数据分别进行建模,对辐照度与功率的概率分布函数进行Copula联合分布函数构建。利用基于经验函数的BFGS(Broy-den,Fletcher,Goldforb and Shanno)参数估计方法对各类Copula函数进行参数估计。参照各Copula函数与经验联合分布函数的欧氏距离与K-S(Kolmogorov-Smirnov)值进行Copula函数选取。结合估计区间最窄原则得出光伏功率条件概率分布90 %置信度下的概率功率曲线。根据工程经验以及考虑异常数据时序特性,依照四类异常数据的判别准则,建立异常数据识别模型。利用光伏电站原始数据与人工合成异常数据进行仿真分析,结果表明该方法能有效、准确地识别出各类异常数据。
Abstract:
      High quality PV power data is the basis for conducting PV research, but the data collected from PV power plants contains a large proportion of abnormal data, so the abnormal data need to be identified. The PV power data of different illumination characteristics were modeled separately, and the Copula joint distribution function was constructed for the probability distribution function of irradiance and power. Using the BFGS (Broy-den, Fletcher, Goldforb and Shanno) parameter estimation method based on empirical function to estimate the parameters of various Copula functions. The selection of Copula function was based on Euclidean distance and K-S (Kolmogorov-Smirnov) values of each Copula function and empirical joint distribution function.The probability power curve under the 90% confidence of the probability distribution of the PV power condition was obtained by combining the narrowest principle of the estimation interval. According to engineering experience and considering the time series characteristics of abnormal data, an abnormal data recognition model was established according to the criteria of four types of abnormal data. The simulation analysis of raw data and synthetic anomaly data of PV power station showed that the method can identify various abnormal data effectively and accurately.
查看全文   查看附录   查看/发表评论  下载PDF阅读器