基于可见光与红外图像的光伏阵列组串热斑检测方法  

GROUP HOT SPOT DETECTION OF PHOTOVOLTAIC ARRAY BASED ON VISIBLE AND INFRARED THERMAL IMAGES

作　　者：黄政之 厉小润[1] 张天文 Huang Zhengzhi;Li Xiaorun;Zhang Tianwen(College of Electrical Engineering,Zhejiang University,Hangzhou 310027,China;Zhejiang Chint New Energy Development Co.,Ltd.,Hangzhou 310052,China)

机构地区：[1]浙江大学电气工程学院,杭州310027 [2]浙江正泰新能源开发有限公司,杭州310052

出　　处：《太阳能学报》2025年第3期454-460,共7页Acta Energiae Solaris Sinica

基　　金：浙江省“尖兵”“领雁”重点研发计划(2023C01129)。

摘　　要：针对光伏阵列组串热斑检测因反光而产生虚警的问题,提出一种基于可见光与红外图像的光伏阵列组串热斑检测方法。首先,考虑到光伏电站的复杂背景易导致虚警,该方法结合UNet神经网络和边缘检测算法完成光伏组件的拆分。根据组串热斑亮度较高的特点,利用聚类算法检测异常高亮区域。最后基于归一化互相关系数模板匹配法实现可见光与红外图像的配准,联合可见光图像的反光区域和光伏组串的温度信息,判定并抑制组串热斑虚警。实验结果表明,该方法适用于多种场景下组串热斑的检测,且能有效消除反光的影响。A group hot spot detection method for PV array strings based on visible and infrared images is proposed aiming at the problem of false alarm due to the sunlight reflection of PV array strings.Firstly,considering that the complex background of PV power plants is easy to cause false alarms,the method combines UNet neural network and edge detection algorithm to realize segmentation and extraction of PV modules.Secondly,based on the characteristic that the gray value of group hot spots is higher than that of normal modules,a clustering algorithm is used to detect abnormally bright areas.Finally,the image alignment is completed by the normalized correlation coefficient template matching method,and the suppression of false alarms is realized according to the temperature information and the reflective region on the visible image.Through experimental comparison and analysis,it is demonstrated that that the proposed method is applicable to the detection of group hotspots in various scenes and can effectively eliminate the influence of sunlight reflections.

关 键 词：光伏组件 红外成像 故障检测 组串热斑 聚类 

分 类 号：TK514[动力工程及工程热物理—热能工程]