摘要
光伏发电受环境温度、天气、湿度等因素影响,具有较强的随机性。本文综合考虑样本数据的数值相似性与形态相似性,提出一种基于FCM-Fréchet的相似日聚类方法。具体步骤如下:首先,基于样本数据的数值相似性,采用FCM算法进行聚类,得到多个聚类集合及其中心;然后,依据数据的形态相似性,利用Fréchet距离对FCM聚类结果进行修正,提高聚类结果的准确性;最后,仿真验证表明,该方法对晴天、阴天、雨天相似日聚类的自相关系数分别为0.959、0.837、0.499,均高于其他传统聚类算法,聚类准确度更高。
Photovoltaic power generation is affected by factors such as ambient temperature,weather,and humidity,exhibiting strong randomness.In this paper,considering both the numerical similarity and morphological similarity of samples,a similar day clustering method based on FCM-Fréchet was proposed.The specific steps are as follows:firstly,a based on the numerical similarity of sample data,the FCM algorithm was adopted for clustering to obtain multiple clustering sets and their centers.Secondly,according to the morphological similarity of the data,the FCM clustering results were modified by Frechet distance to make the clustering results more accurate.Finally,the simulation verification shows that the autocorrelation coefficients of the proposed method for similar day clustering under sunny,cloudy,and rainy days are 0.959,0.837 and 0.499,respectively,all higher than those of other traditional clustering algorithms,indicating superior clustering accuracy.
作者
刘慧杰
刘倩
LIU Huijie;LIU Qian(Guyuan County Power Supply Branch of State Grid Jibei Electric Power Co.,Ltd.,Zhangjiakou 076550,China;State Grid Jibei Electric Power Co.,Ltd.Zhangjiakou Power Supply Company,Zhangjiakou 075000,China)
出处
《电工材料》
2026年第1期102-107,共6页
Electrical Engineering Materials
