风电场异常数据辨识与重构技术综述

Review of Wind Farm Abnormal Data Identification and Reconstruction Techniques

风电场运行数据的质量是影响风电消纳的重要因素,电网调度运行以及电力市场等业务应用都需要风电场数据支撑,但一般情况下,风电场采集到的原始数据会包含部分异常数据。本文详细分析对比了国内外学者对于异常数据辨识和异常数据重构使用的方法。依据数据辨识原理将目前国内外使用的数据辨识方法分为3类,阐述了各辨识方法优越性以及局限性;其次,根据使用重构方法的数量将数据重构方法分为了单一重构法和复合重构法,阐述了各重构法的适用场合以及局限性;然后,根据风电数据的不同应用场合,对异常数据辨识对象以及辨识方法做了详细研究;最后对本领域未来可研究的问题进行了展望。

The quality of wind farm operation data is an important factor affecting wind power consumption.Grid operation and power market operation need wind farm data.However,the raw data collected by wind farms usually contains a large amount of abnormal data.Based on investigation on the sources of anomalous data,this paper analyzes and compares the methods used by scholars for anomalous data identification and anomalous data reconstruction.According to the theories of data identification,the current identification methods currently are divided into three categories.The advantages and limitations of each identification method are expounded.According to the number of reconstruction methods used,the data reconstruction methods are divided into single reconstruction method and composite reconstruction method.Application scenarios and limitations of each reconstruction method are elaborated.In addition,according to the different applications of wind power data,the object of anomalous data identification and the identification method are studied in detail.Finally,the future research issues in this field are presented.