基于CACC-RF的转辙机表示缺口卡阻故障风险预测

CACC-RF-based Risk Prediction of Railway Switch Gap Jam Fault

作为道岔系统的常见故障,转辙机表示缺口卡阻会导致道岔失去位置表示,从而影响线路的通行效率,甚至危及行车安全。基于铁路现场的道岔维护记录,首先利用LS故障风险评价法和k均值聚类实现了对每月转辙机表示缺口卡阻故障发生风险的量化和分级。而后,分析影响每月转辙机表示缺口卡阻故障风险等级的因素,并构建每月的故障风险预测特征向量。最后,基于CACC离散化方法和随机森林,构建转辙机表示缺口卡阻故障风险等级的预测模型,实现对未来各月转辙机表示缺口卡阻故障风险等级的预测。实验结果表明本方法可较为准确地预测各月发生转辙机表示缺口卡阻故障的风险等级,从而指导现场维护人员根据各月不同的故障风险等级提前协调设备、人员和维护天窗等维护资源,调整转辙机表示缺口相关维护活动的强度,最终实现在减少人力物力浪费的同时,提高道岔的维护水平。

As one of the common fault modes in railway turnout system,the railway switch gap jam will cause the loss of position indication of the turnout,which will affect the traffic efficiency of the railway line,and even endanger the safety of the passing trains.Based on the on-site turnout maintenance records,this paper first utilized the LS fault risk assessment method and the k-means clustering to realize the quantification and grading of the monthly risk of the railway switch gap jam fault.Then,the factors that affect the risk level of the switch gap jam fault were analyzed,and the monthly fault risk prediction feature vector was constructed.Finally,based on CACC discretization and random forest,the prediction model of the risk level of the gap jamming fault represented by the switch machine was constructed,and the prediction of the risk level of the switch machine indicating the gap jamming failure in the future months was realized.The experimental results based on the field data show that the method proposed in this paper can accurately predict the monthly risk level of the switch gap jam fault so as to guide the on-site maintenance personnel to coordinate the maintenance resources such as equipment,personnel and maintenance skylight in advance according to the different fault risk level in each month and reasonably adjust the intensity of maintenance activities related to the switch gap,so as to reduce the waste of the manpower and material resources,and improve the maintenance level of the railway turnout.