摘要
铁路信号机房是安装行车设备的关键处所,分布在铁路沿线,点多线长,机房24 h值守需要大量的人力、物力投入。如何采用先进技术解决信号机房日常设备状态远程自动巡视、常见故障远程处理是多年来一直备受关注的课题。为此提出基于5G大带宽、低延时的技术优势,实现设备远程巡检测试、机房环境远程精准监控、作业人员远程管控、应急处置远程辅助等功能;在研究确定基于5G的信号机房智能巡检系统架构设计和设备组成的基础上,分别对高、低分辨率下的视频智能算法进行针对性的遴选、调参和优化,以达到精准识别远程对象的目的。经在怀化西编组站等处所的应用表明:基于5G的信号机房智能巡检系统可以准确识别设备、环境及人员的异常,实现远程日常巡检;由智能巡视机替代人工操作,在提高巡检质量和效率等方面效果明显。
The railway signal equipment room(SER)is a key location for installing train operation equipment,distributed along the railway line with multiple points and long lines.The 24-hour duty of the SERs requires high manpower and material resources input.How to adopt advanced technology to solve the remote automatic patrol of daily equipment status and remote handling of common faults in SERs has been a topic of great concern for many years.To this end,based on the advantages of high bandwidth and low latency of 5G technology, remote patrol and testing of equipment, remote precise monitoring of the SER environment, remote control of SER operators, and remote assistance for emergency response are proposed. The architecture design and equipment composition of an intelligent patrol system for SERs based on 5G technology have been determined through research. Targeted selection, parameter tuning, and optimization were carried out on video intelligent algorithms for high-resolution and low-resolution, respectively, to achieve accurate recognition of remote objects. The application in Huaihua West Marshalling Yard and other places shows that the intelligent patrol system for SERs based on 5G technology can accurately identify abnormalities in equipment, environment, and personnel, and has achieved remote daily patrol. The intelligent patrol machine replaces manual operation, which has obvious effect in improving patrol quality and efficiency.
作者
钟志旺
林建国
ZHONG Zhiwang;LIN Jianguo
出处
《铁道通信信号》
2024年第11期9-14,共6页
Railway Signalling & Communication
基金
中国铁路广州局集团有限公司科技研究开发计划(2022K155-K)。
关键词
智能巡检
5G技术
信号机房
远程管控
应急处置
Intelligent patrol
5G technology
Signal equipment room
Remote control
Emergency response
