基于改进YOLOv5的巡更安全风险识别方法研究  

Research on Patrol Safety Risk Identification Method Based on Improved YOLO v5

作　　者：齐芳平 石晔 崔志威 王辉 朱澈 QI Fangping;SHI Ye;CUI Zhiwei;WANG Hui;ZHU Che(Huaihe Energy Group Coalbed Methane Development and Utilization Company)

机构地区：[1]淮河能源集团煤层气开发利用公司

出　　处：《上海节能》2023年第12期1876-1881,共6页Shanghai Energy Saving

摘　　要：为了自动检测巡更人员是否佩戴安全帽,提高单人巡检的安全性,提出一种针对瓦斯电站内巡更人员的安全风险识别方法。首先使用网络爬虫收集佩戴安全帽的图像构建数据集,其次使用YOLO v5对构建好的数据集进行训练,训练出安全帽检测模型,最后在实际生产过程中进行应用。该算法通过YOLO v5模型对巡更人员是否佩戴安全帽进行检测,并记录人员在厂房内的逗留时长,在超时后及时发出警报,有效保护了巡更人员的生命安全。同时,依据算法检测是否佩戴安全帽,减少了一定的人力输出,更具有一定的经济效应。In order to automatically detect whether the patrol personnel are wearing safety helmets and improve the safety of individual patrol inspection,a safety risk identification method for patrol personnel in gas power stations is proposed.Firstly,a data set is constructed by using a web crawler to collect im-ages of people wearing safety helmets.Secondly,YOLO v5 is used to train the constructed data set to generate a safety helmet detection model.Finally,the algorithm is applied in the actual production pro-cess.The algorithm detects whether the patrol personnel are wearing safety helmets through the YO-LO v5 model,records the length of stay in the plant,and sends an alarm in time after timeout,effective-ly protecting the life safety of the patrol personnel.At the same time,based on the algorithm to detect whether to wear safety helmets,it reduces certain human output and has certain economic effects.

关 键 词：瓦斯电站 安全帽 YOLO v5 安全防护 经济性 

分 类 号：TD712.67[矿业工程—矿井通风与安全] TM619[电气工程—电力系统及自动化] TP391.41[自动化与计算机技术—计算机应用技术]