摘要
当前油气站场对作业人员安全帽佩戴情况的监管主要依赖于监控摄像头的视频排查和人工实地巡检,缺少智能监测、识别、分析及报警的先进功能。针对上述问题,设计了油气站场安全帽佩戴检测系统。该系统采用了YOLOv5s模型作为油气站场人员安全帽佩戴检测算法,训练后的YOLOv5s模型的平均精度达到了93.62%。该系统还可与油气站场现有智能化系统进行集成与联动,实现信息的互联互通和站场的协同管理。应用表明,该系统能够准确识别未正确佩戴安全帽的作业人员,提高了油气站场智能化安全管理能力,确保了油气田生产的安全高效运行。
Currently,the supervision of safety helmet wearing by personnel in oil and gas stations mainly relies on video screening by surveillance cameras and on-site manual inspections,which is lack of advanced functions such as intelligent monitoring,recognition,analysis,and alarm.In response to above problems,a safety helmet wearing detection system for oil and gas stations is designed.The YOLOV5s model is adopted as the algorithm for detecting safety helmet wearing by personnel,the average precision of the trained YOLOv5s model reaches 93.62%.The system can also be integrated and linked with the existing intelligent systems in oil and gas stations to achieve information interconnection and interoperability and collaborative management of the stations.Applications show that the system can accurately identify personnel who have not correctly worn safety helmets.The intelligent safety management capabilities of oil and gas stations is improved,the safe and efficient operation of oil and gas fields are guaranteed.
作者
朱威铭
Zhu Weiming(PipeChina Engineering Technology Innovation Co.Ltd.,Tianjin,300000,China)
出处
《石油化工自动化》
2025年第3期52-56,61,共6页
Automation in Petro-chemical Industry
关键词
油气站场
安全帽佩戴检测
目标检测算法
YOLOv5s模型
站场系统集成
oil and gas station
safety helmet wearing detection
object detection algorithm
YOLOv5s model
station system integration
