###
DOI:
电力大数据:2023,26(10):-
←前一篇   |   后一篇→
本文二维码信息
基于机器学习算法的光伏组串故障诊断技术研究
(中国电力工程顾问集团中南电力设计院有限公司)
Research on Fault Diagnosis Technology for Photovoltaic String Based on Machine Learning Algorithms
Luo Yuan Peng1,2, Fu JiangQue1,2, Li HongMing1,2, Zhang Qi1,2, Zhang WenCheng1,2
(1.Central Southern China Electric Power Design Institute Co.,Ltd,of China Power Engineering Consulting Group,Wuhan Hubei 430000;2.China)
摘要
图/表
参考文献
相似文献
本文已被:浏览 240次   下载 793
投稿时间:2023-05-30    修订日期:2023-08-19
中文摘要: 常规缺陷检测方法中,主要依据光伏电站面板异常状态数据检测面板缺陷,检测结果存在着一定的随机性,导致缺陷检测结果不清晰。因此,利用了无人机影像技术,设计了光伏电站面板缺陷检测方法。提取出图像中的缺陷特征,结合无人机影像技术,通过灰度共生矩阵将缺陷图像与完整图像分割开来,识别可见光图像缺陷位置,并将缺陷图像放在光伏面板缺陷检测模型中进一步检测,使图像纹理特征与形状特征高度融合,从而实现光伏电站面板缺陷的精准检测。采用对比实验的方式,验证了该检测方法的检测置信度更高,检测精准度随之升高,能够应用于实际生活中。
中文关键词: 无人机影像技术  光伏电站  面板  缺陷  检测方法  
Abstract:In the conventional defect detection methods, panel defects are mainly detected based on the abnormal status data of the photovoltaic power plant panel. The detection results have certain randomness, resulting in unclear defect detection results. Therefore, the UAV image technology is used to design a defect detection method for photovoltaic power plant panel. The defect features in the image are extracted, combined with the UAV image technology, the defect image is separated from the complete image through the gray level co-occurrence matrix, the defect location of the visible light image is recognized, and the defect image is placed in the photovoltaic panel defect detection model for further detection, so that the image texture features and shape features are highly integrated, so as to achieve the accurate detection of photovoltaic power plant panel defects. By means of comparative experiments, it is verified that the detection method has higher detection confidence and detection accuracy, which can be applied in real life.
文章编号:     中图分类号:    文献标志码:
基金项目:
引用文本: