一种光伏阵列故障识别与定位系统  

A Photovoltaic Array Fault Identification and Location System

作　　者：周颖 王小燕[1] Zhou Ying;Wang Xiao-yan(Anhui Industry&Trade Vocational Technical College,Anhui Huainan 232009)

机构地区：[1]安徽工贸职业技术学院,安徽淮南232009

出　　处：《电子质量》2022年第9期67-69,共3页Electronics Quality

基　　金：安徽高校自然科学研究项目(KJ2021A1343)资助。

摘　　要：针对目前光伏阵列检测巡检时间长、效率低等缺点,一种利用无人巡检识别故障并定位的系统被设计出来。该系统采用图像处理技术与特征提取方法,使用无人机和无线数传模块等硬件实现了对光伏阵列光伏板的快速巡检。软件编写采用OTSU算法进行视频图像截取、图像SIFT特征提取。然后对提取到SIFT特征的图像建立Bag of Word模型,使用支持向量机(SVM)计算图像模型得分,根据分值最后甄别出故障类型图像。通过时间、飞行速率、GPS和标志物等信息,对故障光伏板进行定位。经实测,故障识别正确率超过80%。Aiming at the disadvantages of long inspection time and low efficiency of photovoltaic array detection, a system for fault identification and location using unmanned inspection is designed. The system adopts image processing technology and feature extraction method, and uses hardware such as UAV and wireless data transmission module to realize rapid inspection of photovoltaic array photovoltaic panels. The software adopts OTSU algorithm for video image capture and image SIFT feature extraction. Then, the bag of word model is established for the image with SIFT features extracted, and the support vector machine(SVM) is used to calculate the score of the image model. Finally, the fault type image is identified according to the score. The fault photovoltaic panel can be located through time, flight speed, GPS and markers. The actual measurement shows that the correct rate of fault identification is over 80%.

关 键 词：视频传输 图像截取 特征量提取 故障定位 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置] TM615[自动化与计算机技术—控制科学与工程]