地铁列车碳滑板磨损区域及磨耗预测  

Prediction on Wear Area and Wear of Carbon Slider in Subway Train

作　　者：阮海祺 雷斌[1,3,4] 杨克臣 RUAN Haiqi;LEI Bin;YANG Kechen(Institute of Mechanical and Electrical Technology,Lanzhou Jiaotong University,Lanzhou 730070,China;Shenzhen Metro Group Co.,Ltd.,Shenzhen 518040,China;Gansu Logistics and Transportation Equipment Information Technology Research Center,Lanzhou 730070,China;Gansu Logistics and Transportation Equipment Industry Technical Center,Lanzhou 730070,China)

机构地区：[1]兰州交通大学机电技术研究所,甘肃兰州730070 [2]深圳地铁运营集团有限公司,广东深圳518040 [3]甘肃省物流及运输装备信息化工程技术研究中心,甘肃兰州730070 [4]甘肃省物流与运输装备行业技术中心,甘肃兰州730070

出　　处：《铁道车辆》2025年第1期102-108,共7页Rolling Stock

基　　金：国家自然科学基金(72061021);甘肃省自然科学基金(21JR7RA284)。

摘　　要：在对城市轨道交通车辆的检修工作进行深入调研的过程中,发现部分受电弓碳滑板存在较为严重的磨损现象,这一问题对弓网结合度以及列车安全运行产生了较大影响。为了确保线路上新碳滑板的正常运维,有必要准确预测受电弓碳滑板的磨损量及其变化趋势。在智慧运维系统的快速发展背景下,文章综合考虑了传统统计数据,并引入长短期记忆网络(Long Short-Term Memory,LSTM)进行了受电弓碳滑板的磨耗区域及磨耗量的预测。通过应用不同的预测算法模型进行对比分析,实验结果显示,LSTM网络在碳滑板磨耗预测方面的精度较其他机器学习模型高出约10%,这一结果表明,LSTM网络在预测多个更换周期内碳滑板磨损方面具有显著优势。具体而言,文章通过建立和训练LSTM模型,对碳滑板的磨损数据进行了系统化的分析和预测,对比结果表明,LSTM模型不仅能够准确捕捉磨损趋势,而且其预测数据与实际实验结果高度一致,基本满足实际使用的需求,可为城市轨道交通车辆的维护和管理提供有力的技术支持,能够有效提高运营安全性并延长碳滑板的使用寿命。During the in-depth investigation of the maintenance of urban rail vehicles,it was found that the carbon slider of pantographs had serious wear problems,which had a significant impact on the integration of pantograph and catenary and the safe operation of the train.In order to ensure the normal operation of the new carbon sliders,it is necessary to accurately predict the wear amount and changing trend of carbon sliders.In the context of the rapid development of intelligent operation and maintenance systems,this paper comprehensively considers traditional statistical data and introduces Long Short-Term Memory network(LSTM)to predict the wear area and wear amount of carbon sliders.By applying different prediction algorithm models for comparative analysis,the test results show that the accuracy of LSTM network in predicting wear of carbon sliders is about 10%higher than other machine learning models,indicating that LSTM network has significant advantages in predicting the wear of carbon sliders in multiple replacement cycles.Specifically,this paper systematically analyzes and predicts the wear data of carbon sliders by establishing and training LSTM model.The comparison results show that the LSTM model can not only accurately capture the wear trend,but also its predicted data are highly consistent with the actual test results,which basically meets the needs of actual use.It can provide strong technical support for the maintenance and management of urban rail vehicles,effectively improve the operation safety,and prolong the service life of carbon sliders.

关 键 词：地铁列车 智慧运维 碳滑板磨耗预测 变化趋势 长短期记忆网络 安全性 

分 类 号：U264.34[机械工程—车辆工程]