基于改进C4.5算法的新型车辆故障预测方法研究  

Research on New Vehicle Fault Prediction Method Based on Improved C4.5 Algorithm

作　　者：戴银娟 付石磊 DAI Yinjuan;FU Shilei(Mechatronics T&R Institute, Lanzhou Jiaotong University, Lanzhou 730070;Engineering Technology Center for Information of Logistics & Transport Equipment, 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]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

出　　处：《常熟理工学院学报》2019年第5期72-77,共6页Journal of Changshu Institute of Technology

摘　　要：无接触网供电车辆结构复杂,设备状态监测系统采集到的数据量大,数据类型多,价值密度低.为了充分挖掘其潜在的信息,传统的数据处理通过自身经验发现数据中的价值,预测精度较低,不能满足目前的需求. C4.5决策树算法作为数据挖掘中常用的数据处理方式,具有一定的优越性.本文基于改进的C4.5算法,首先在传统算法的基础上,提出一种改进的决策树分类算法;再基于该算法流程建立了无接触网供电新型车辆故障预测模型;最后通过案例验证了预测的准确性,为列车安全可靠运行提供了有效保障措施.The structure of the contactless network power supply vehicle is complex. The equipment state monitoring system collects a large amount of data which has many data types and low value density. In order to fully exploit its potential information, the traditional data processing has its drawbacks and cannot meet the current needs. There are certain advantages in the data classification methods commonly used in decision tree data mining. Based on the improved C4.5 algorithm, this paper first proposes an improved decision tree classification algorithm based on the traditional algorithm, and then builds a contactless network-powered urban rail vehicle fault prediction model based on the improved C4.5 decision tree algorithm flow. The case verifies the accuracy of its prediction, which provides an effective guarantee for the safe and reliable operation of the train and reduces the occurrence of faults.

关 键 词：无接触网供电 C4.5 决策树 故障预测 

分 类 号：U223.6[交通运输工程—道路与铁道工程]