Wind power data cleaning method based on isolated forest and local outlier factor

• Main Authors: GE XUEYANG, YAN YUTONG, ZHANG SHUYU, PENG XIAOYU, CHEN QI, LIU SHENCHENG, LIU HONGBO, ZHANG CHONG, LIU YONGFA
• Format: Patent
• Language: Chinese; English
• Published: 04.01.2022
• Subjects: CALCULATING; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; HANDLING RECORD CARRIERS; PHYSICS; PRESENTATION OF DATA; RECOGNITION OF DATA; RECORD CARRIERS

The invention discloses a wind power data cleaning method based on an isolated forest and a local outlier factor. Based on abnormal data identification of an isolated forest and a local outlier factor and data reconstruction of wind power and related data thereof, data cleaning work of the wind power and the related data thereof is completed; the problem of cleaning abnormal data generated by traditional wind power and related data or abnormal data caused by abnormal weather in the collection and transmission process is effectively solved. In the process of reconstructing missing values in the data, a proper interpolation algorithm is adopted according to different data types, so that the integrity of the data is ensured; and meanwhile, the reasonability and the time sequence of the data are ensured, and the invention plays a more efficient role in subsequent wind power prediction or other purposes. 本发明一种基于孤立森林及局部离群因子的风功率数据清洗方法，基于孤立森林及局部离群因子的异常数据识别和风功率及其相关数据的数据重构，完成了风功率及其相关数据的数据清洗工作，有效解决了在采集、传输过程中，传统风电功率及其相关数