摘要
在满足日益增多的可再生能源需求的情况下,光伏发电系统的广泛部署和高效运行变得必不可少。但是,光伏发电系统的故障不仅对能源产出产生影响,而且可能增加维护成本和系统停机时间。传统的故障检测方式往往依赖人工巡检,既费时,效率又不高。因此,探索人工智能技术在光伏组件故障检测与维护方面的应用就变得尤为重要。为此,提出了一种自动化的光伏组件故障检测方法,其基础是对光伏组件进行高分辨率成像,然后利用深度学习算法分析图像数据,以识别并分类不同类型的故障,如热斑裂纹和遮挡等。开发了一套以AI为基础的维护决策支持系统,可基于故障严重程度和组件位置智能规划维护策略,实验结果证明所提出的AI驱动方法在故障检测上表现出高精确率和快速反应能力,使光伏系统的运维效率得到显著提高。此项研究成果不仅为光伏行业提供了技术支持,还为今后智能光伏系统的开发和优化奠定了基础。
In order to meet the increasing demand for renewable energy,the widespread deployment and efficient operation of photovoltaic power generation systems have become essential.However,the failure of photovoltaic power generation systems not only affects energy output,but may also increase maintenance costs and system downtime.Traditional fault detection methods often rely on manual inspection,which is both time-consuming and inefficient.Therefore,exploring the application of artificial intelligence technology in photovoltaic module fault detection and maintenance has become particularly important.This article proposes an automated fault detection method for photovoltaic modules,which is based on high-resolution imaging of photovoltaic modules,and then uses deep learning algorithms to analyze image data to identify and classify different types of faults,such as hot spot cracks and occlusion.We have also developed an AI based maintenance decision support system that can intelligently plan maintenance strategies based on fault severity and component location.Experimental results have shown that the proposed AI driven method exhibits high accuracy and rapid response ability in fault detection,significantly improving the operational efficiency of photovoltaic systems.This research result not only provides technological support for the photovoltaic industry,but also lays the foundation for the development and optimization of intelligent photovoltaic systems in the future.
作者
胡伟
王迪
胡耀蓉
冯钰玮
赵柯
HU Wei;WANG Di;HU Yaorong;FENG Yuwei;ZHAO Ke(PowerChina Guizhou Electric Power Engineering Co.,Ltd.,Guiyang 550000,China)
出处
《电工技术》
2024年第24期39-43,49,共6页
Electric Engineering
基金
贵州院专项资金资助“贵州省可再生能源数字化应用关键技术研究及平台开发”(编号GZEDKJ-2023-05)。
关键词
人工智能
光伏组件
故障检测
维护策略
机器学习
artificial intelligence
photovoltaic modules
fault detection
maintenance strategies
machine learning
