铁路周界入侵监测告警系统设计与实现  

Design and Implementation of Railway Perimeter Intrusion Monitoring and Alarm System

作　　者：穆振华 MU Zhenhua(Guoneng Baoshen Railway Group Co.,Ltd.,Baotou Inner Mongolia 014010)

机构地区：[1]国能包神铁路集团有限责任公司,内蒙古包头014010

出　　处：《中国科技纵横》2024年第16期55-58,共4页China Science & Technology Overview

摘　　要：铁路是重要的能源运输工具,现有人工巡线在异物入侵检测上存在较大的安全隐患,列车行车安全是需要重点研究的课题。本文利用智能感知技术,辅助现场人工智能巡查和护路巡线,可以实现翻越护栏、横穿股道、人车识别等场景的智能感知。通过YOLOv5算法和边缘计算终端实现视频数据的即时提取和解析,大幅提高铁路周界入侵预警信息的即时性和可靠性。同时,利用雷视一体机实现视频图像数据的高清抓取,提高系统运行稳定性,拓宽边缘计算技术在重载铁路的应用,有效保障铁路列车行车安全。As an important means of energy transportation,railways,the existing manual patrols pose significant security risks in foreign object intrusion detection,train operation safety is a key issue that needs to be studied.This article utilizes intelligent perception technology to,assisting on-site artificial intelligence inspections and road patrol,intelligent perception can be achieved in scenarios such as crossing guardrails,crossing tracks,and recognizing pedestrians and vehicles,real time extraction and analysis of video data are realized through YOLOv5 algorithm and edge computing terminal,significantly improving the timeliness and reliability of railway perimeter intrusion warning information,at the same time,high-definition capture of video image data is achieved through the Thunder Vision all-in-one machine,improve the stability of system operation,widened the application of edge computing technology in heavy haul railway,it can effectively ensure the safety of railway train operation.

关 键 词：异物入侵检测 边缘计算 YOLOv5算法 铁路周界 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]