机务安全监测及检测数据融合与应用研究  

Research on Fusion of Locomotive Safety Monitoring and Detection Data and Its Application

作　　者：龚利 谯兵 赵超 姚凯 GONG Li;QIAO Bin;ZHAO Chao;YAO Kai(Institute of Electronic Computing Technology,China Academy of Railway Sciences Corporation Limited,Beijing 100081,China;Beijing Jingwei Information Technology Co.,Ltd.,Beijing 100032,China;Zhuzhou CRRC Times Electric Co.,Ltd.,Zhuzhou,Hunan 412001,China;Chengdu Yunda Technology Co.,Ltd.,Chengdu,Sichuan 611731,China)

机构地区：[1]中国铁道科学研究院集团有限公司电子计算技术研究所,北京100081 [2]北京经纬信息技术有限公司,北京100032 [3]株洲中车时代电气股份有限公司,湖南株洲412001 [4]成都运达科技股份有限公司,四川成都611731

出　　处：《控制与信息技术》2020年第5期62-67,共6页CONTROL AND INFORMATION TECHNOLOGY

基　　金：中国铁路总公司科技研究开发计划(N2018J009)。

摘　　要：机务安全监测检测数据是中国铁路机车运行时的重要信息,对其进行分析诊断及评估预测对保障机车运行安全有重要作用。文章提出了一套利用基于任务流的ETL数据抽取算法实现机车安全监测检测数据集成与融合的方案,其针对机车次生故障问题,提出清洗方法,实现了智能专家分析;针对传统机务维修维护方式存在的周期长、成本高、效率低等问题,提出基于决策树算法的设备质量诊断、评估预测方法,并构建了智能机务维修维护体系;通过搭建基于大数据的机务安全监测检测数据平台,实现了机车故障诊断的远程化、智能化与协同化,为机车的安全稳定运行提供保障。Data of locomotive safety monitoring and detection is an important running information of Chinese railway locomotive.This paper introduced a scheme of integration and fusion of locomotive safety monitoring and detection data by a task flow based ETL data extraction algorithm.A cleaning scheme for locomotive secondary failure is put forward and the intelligent expert analysis is realized.To solve the problems of long cycle,high cost and low efficiency of traditional maintenance methods,a method of equipment quality diagnosis,evaluation and prediction based on decision tree algorithm is put forward to build an intelligent maintenance system.A big data based data platform for locomotive safety monitoring and detection is built to realize remote,intelligent and collaborative fault diagnosis of locomotive,so as to guarantee the safe and stable operation of locomotive.

关 键 词：机务安全监测检测 大数据 数据融合 智能专家分析 

分 类 号：TP311.5[自动化与计算机技术—计算机软件与理论]