基于人工智能技术的电气化铁路接触网检测研究  

Research on the Detection of Electrified Railway Contact Network Based on Artificial Intelligence Technology

作　　者：付宝宝 Fu Baobao(China Railway Construction Electrification Group North Company,Taiyuan,Shanxi,China,030000)

机构地区：[1]中国铁建电气化集团北方公司,山西太原030000

出　　处：《仪器仪表用户》2025年第1期15-17,20,共4页Instrumentation

摘　　要：传统的铁路接触网检测工作依赖人工肉眼观察,存在工作量大、效率低、精度差等问题,必须积极探寻一种更为科学高效的检测方法。本文基于人工智能技术,对电气化铁路接触网检测技术进行研究,设计出基于双目视觉的电气化铁路接触网检测系统,使用一种结构化多分辨率深度图像采集和处理算法,通过双目摄像机获取接触网图像,再通过DCNN网络进行目标检测与定位,从而实现对接触网接触线、承力索、绝缘子等关键部件的状态评估及故障识别,解决了现有技术无法解决的诸多难题,提高了巡检效率,降低了作业成本,为进一步提升电气化铁路安全性能提供有力保障。Traditional railway catenary inspection relies heavily on manual visual inspection,which is characterized by heavy workloads,low efficiency,and poor accuracy.Therefore,it is imperative to explore a more scientific and efficient inspection method.This article investigates the technology of electrified railway catenary inspection based on artificial intelligence and designs a binocular vision-based inspection system for electrified railway catenaries.This system employs a structured multi-resolution deep image acquisition and processing algorithm.By using binocular cameras to capture catenary images,and then utilizing a Deep Convolutional Neural Network(DCNN)for target detection and localization,it assesses the condition and identifies faults in key components such as contact wires,messenger wires,and insulators.This approach resolves numerous challenges that cannot be tackled by existing technologies,improving inspection efficiency,reducing operational costs,and providing a solid foundation for enhancing the safety performance of electrified railways.

关 键 词：人工智能 铁路接触网检测 电气化 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]