基于深度学习的铁路货场作业安全状态识别与监控技术研究

Identification and Monitoring of Operation Safety Status in Railway Freight Yard Based on Deep Learning

近年来,我国铁路已在管理信息化、货运数字化等方面取得了一定的成就,但货运全过程安全问题的智能化监控、识别、分析、预警能力有所欠缺。铁路货场作业的监督仍然借助前端摄像机,同时依赖人工盯控、经验决策为主的粗放式安防作业模式识别货场作业安全风险,工作强度大,容易形成安全隐患。通过对铁路货场作业风险识别项点进行分析,提取出静态不安全状态和动态不安全状态识别项点;采用深度学习技术构建了从智能识别到自动警示的铁路货场作业安全监管系统框架,提出实现铁路货场安全作业的技术解决思路,对保障人身安全、提升货物运输安全监测及预警水平具有重要意义。

In recent years,China’s railway has made certain achievements in management informatization,freight transportation digitalization,and other aspects.However,the whole process of freight transport still lacks intelligent monitoring,identification,analysis,and early warning of safety issues.Railway freight yard operations are still supervised with the help of front-end cameras,and the identification of safety risks during operations relies on an extensive mode featuring manual control and decisions from experience.Safety hazards tend to occur under this mode with high work intensity.This paper extracted the points of static unsafety status and dynamic unsafety status identification by analyzing the points of operation safety risks identification in railway freight yards.Then,the deep learning technology was utilized to construct the framework of the operation safety supervision system in railway freight yards encompassing a whole process from intelligent identification to automatic warning.Finally,the technical solution ideas for realizing safe operation in railway freight yards were proposed.This achievement is of great significance in ensuring personal safety and improving safety monitoring and early warning in freight transport.