摘要
高速铁路工务设备是铁路运输的基础,其使用状态直接关系到运输的安全性。为提升工务段巡检员和管理层的巡检效率和决策的准确性,解决巡检预警状态缺乏直观性、周期天数设置不合理以及春秋季专项检查缺乏电子化等问题,本文研究了两项关键技术,以构建数字化技术体系。首先利用深度学习建立了基于卷积神经网络的病害自动识别模型,并通过对几个高速铁路沿线车间的数据进行采集与验证分析,证明该模型在巡检员上传病害影像后,能够自动识别病害。训练精度高达97%,验证精度达到76%。其次,本文基于大数据技术,结合卷积神经网络和长短期神经网络建立设备状态判断模型,并构建出巡检周期预测算法。通过分析两个高速铁路工务段的数据,结果显示,设备状态识别模型能够通过捕捉巡检记录中的关键信息,准确判断设备运行状态,预测巡检次数的平均相对误差为14.3%,精度为85.7%。最后,通过高速铁路工务设备巡检修数字化实现案例,充分证明了两项应用的设计与融合方案能够为高速铁路工务设备的巡检修作业提供充分的技术支持,并能为实时而高效的检修决策提供全面可靠的数据依据。
High-speed railway engineering equipment is the basis of railway transportation,and its use status is directly related to the safety of transportation.In order to improve the inspection efficiency and decision-making accuracy of inspectors and managers of the engineering departments,solve the problems such as lack of intuitiveness of inspection early warning status,unreasonable cycle days setting and lack of electronization of spring and autumn special inspection,and realize intelligent and digital management,this paper studies two key technologies to build a digital technology system.Firstly,deep learning was used to establish an automatic disease recognition model based on convolutional neural network.Through data collection and verification analysis of workshops along several high-speed railways,it is proved that the model is able to automatically identify diseases after uploading disease images.The training accuracy is up to 97%,and the verification accuracy is up to 76%.Secondly,based on big data technology,this paper combines convolutional neural network and long-short term memory to establish a equipment state judgment model,and builds an inspection frequency prediction algorithm.By analyzing the data of two high-speed railway work sections,the results show that the equipment status judgment model is able to accurately judge the equipment operating status by capturing the key information in the inspection records,and the average relative error of predicting the inspection times is 14.3%,and the accuracy is 85.7%.Finally,through the case of digital realization of inspection and repair of high-speed railway engineering equipment,it is fully proved that the design and fusion of these two applications can provide sufficient technical support for the inspection and repair of high-speed railway engineering equipment,and can provide comprehensive and reliable data basis for real-time and efficient inspection and repair decisions.
作者
周小爱
孙瑞海
代冲
肖翔
Zhou Xiaoai;Sun Ruihai;Dai Chong;Xiao Xiang(Institute of Computing Technology,China Academy of Railway Sciences,Beijing 100081,China;Beijing-Shanghai High Speed Railway Co.,Ltd.,Beijing 100038,China;Key Laboratory of Railway Industry for Operational Active Safety Assurance and Risk Prevention and Control,Beijing Jiaotong University,Beijing 100044,China)
出处
《铁道技术标准（中英文）》
2024年第6期25-34,共10页
Railway Technical Standard(Chinese & English)
基金
国家重点研发计划(2022YFB4300605)。
关键词
高速铁路
工务设备
巡检修
卷积神经网络
病害自动设别模型
长短期记忆网络
high-speed railway
engineering equipment
inspection and repair
convolutional neural network
automatic disease recognition model
long short-term memory
