YOLO系列算法在电力行业目标检测领域的应用与发展趋势  

Application and Development Trends of YOLO Series Algorithms in the Field of Object Detection in the Power Industry

作　　者：张豪 高林 龚宇翔 伏德粟 ZHANG Hao;GAO Lin;GONG Yuxiang;FU Desu(College of Intelligent Systems Science and Engineering,Hubei Minzu University,Enshi 445000,China;Zhejiang Longyuan New Energy Development Co.,Ltd.,Hangzhou 310000,China)

机构地区：[1]湖北民族大学智能科学与工程学院,湖北恩施445000 [2]浙江龙源新能源发展有限公司,杭州310000

出　　处：《湖北民族大学学报(自然科学版)》2025年第1期86-93,共8页Journal of Hubei Minzu University:Natural Science Edition

基　　金：国家自然科学基金项目(12464004,61562025);湖北省高等学校省级教学研究项目(2017387);湖北民族大学校内科研项目(XN2317)。

摘　　要：为了研究你只看一次(you only look once, YOLO)系列算法在电力行业目标检测领域的应用情况,分析其未来在该行业的发展趋势。首先分析了较新的YOLO版本10(YOLO version 10,YOLOv10)算法的网络结构,然后探讨了YOLO系列算法在电力行业从发电、输电、变电到用电环节中目标检测的应用,最后从潜在改进方向、与大模型的融合2方面分析了YOLO系列算法的发展趋势。研究发现,YOLO系列算法在检测速度和精度方面取得了明显进展,特别是在电力行业的缺陷检测、故障检测、设备监控、智慧管理、安全监测等方面表现出极大的潜力;但在复杂背景下,该系列算法仍存在检测精度不高的问题。YOLO系列算法要在电力行业中实现更广泛的应用,还需进一步优化算法的速度与精度以应对实际应用中的挑战。In order to study the application of the you only look once(YOLO)series algorithms in the field of power industry object detection,and analyze their future development trends in the industry.First,the new YOLO version 10(YOLOv10)algorithm of network structure was analyzed.Second,the application of the YOLO series algorithms to object detection in the power industry from power generation,transmission,transformation to utilization were discussed.Finally,the development trends of the YOLO series algorithms were analyzed from two aspects,namely,the potential improvement direction and the fusion of large model.The study found that significant progress had been made in terms of detection speed and precision,with substantial potential demonstrated in defect detection,fault detection,equipment monitoring,intelligent management and security monitoring.However,in the complex background,the series of algorithms still had problem of poor detection precision.The YOLO series algorithms need to be gradually applied more widely in the power industry,and further optimizations in speed and precision will be needed to address practical challenges.

关 键 词：人工智能 YOLO系列算法 电力行业 目标检测 改进模型 未来趋势 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]