山东电力技术2026，Vol.53Issue(6)：13-22,10.DOI:10.20097/j.cnki.issn1007-9904.250305
基于天气分型聚类和特征参数提取的分布式光伏识别方法
A Distributed Photovoltaic Identification Method Based on Weather Type Clustering and Feature Parameter Extraction
摘要
Abstract
In recent years,with the continuous development of new energy generation technology,distributed photovoltaics(DPVs)have gradually become popular among residents.However,a large number of"unmonitored"DPVs pose great challenges to the safe and stable operation of the power grid.In order to improve the identification accuracy of DPVs,a method based on weather type clustering and the extraction of net load curve feature parameters is proposed in this paper.Firstly,K-means clustering is used to cluster and analyze the net load data of users according to four different weather conditions,and the net load curves for the corresponding weather conditions are obtained.Then,characteristic parameters of the net load curves are extracted,and the random forest algorithm is employed to establish a DPV identification model.Finally,by integrating various characteristic parameters and evaluation indicators to determine whether the user has installed distributed photovoltaics,it can be concluded that the accuracy of the proposed method is higher than that of other identification methods.关键词
天气分型/用户净负荷/K-means聚类/特征参数提取/随机森林Key words
weather type/customer net load/K-means clustering/feature parameter extraction/random forest分类
信息技术与安全科学引用本文复制引用
孟政吉,汪子健,王磊,胡雪凯,史林军..基于天气分型聚类和特征参数提取的分布式光伏识别方法[J].山东电力技术,2026,53(6):13-22,10.基金项目
国家电网有限公司科技项目(kj2024-027).Science and Technology Project of State Grid Corporation of China(kj2024-027). （kj2024-027）
