摘要
列车运行过程中的突发故障常常会给铁路运输系统带来严重影响,导致列车延误、运行计划紊乱等问题。通过对铁路故障数据进行分析和应用,能为调度指挥决策提供有效信息,提升调度指挥决策效率和效用。首先,提出铁路故障数据的文本数据预处理方法,实现分词和数据清洗;进而,基于卷积神经网络的文本处理方法构建故障分类模型,以提高故障分类准确度和均衡性,并通过历史数据分析,选择分布拟合函数并确定拟合参数,实现对故障持续时间预测;同时,构建逻辑模型,考虑故障发生区间和相邻区间,通过列车在站内或区间的增加晚点情况判定列车运行过程是否受故障影响;最后,通过案例验证上述方法的可行性。
Sudden failures in train operation often bring serious impact on railway transport system,resulting in problems such as train delay and operation plan disruption.The analysis and application of these fault data can provide effective information for dispatching&commanding decision-making,and effectively improve the efficiency and effectiveness of dispatching&commanding decision-making.This paper first proposes a text data pre-processing method of railway fault data to realize word segmentation and data cleaning;then it builds a fault classification model based on the text processing method of convolutional neural network to improve accuracy and balance of fault classification.Through historical data analysis,it selects distribution fitting functions and determines fitting parameters to predict the fault duration.At the same time,a logical model is built to cover the fault section and adjacent sections,and to determine whether the train operation is affected by the fault through the increase of train delay at station or in section.Finally,the papier verifies the feasibility of the above method through case study.
作者
洪鑫
江世玉
张红斌
杜何伟
陈亚茹
HONG Xin;JIANG Shiyu;ZHANG Hongbin;DU Hewei;CHEN Yaru(Institute of Computing Technology,China Academy of Railway Sciences Corporation Limited,Beijing 100081,China;School of Traffic and Transportation,Beijing Jiaotong University,Beijing 100044,China;Beijing Jingwei Information Technology Co.,Ltd.,Beijing 100081,China)
出处
《中国铁路》
北大核心
2024年第12期15-23,共9页
China Railway
基金
中国国家铁路集团有限公司科技研究开发计划项目(P2023X012)
中国铁道科学研究院集团有限公司科研开发基金项目(2022YJ356)
关键词
铁路运输
调度指挥
故障文本分析
故障分类
故障持续时间预测
故障影响识别
railway transport
dispatching command
fault text analysis
fault classification
fault duration prediction
fault impact identification
