地铁列车常用气制动故障特征提取

Fault Feature Extraction of Service Air Braking of Subway Train

摘要:

为了对地铁列车常用气制动的隐患进行挖掘，选择中继阀前端气缸容积（CV）实际压力与目标压力的跟随性能作为故障特征；利用地铁列车按模式运行的特点，根据CV目标压力变化情况，将常用气制动过程划分为减速阶段、停车保持阶段和缓解阶段等三个典型阶段.根据各阶段特点，提出了三种故障特征提取方法：相关系数法、平均绝对误差法和直接提取法，得到了6个故障特征.最后，在制动系统可靠性试验台上试验模拟了三类典型隐患故障，采用对故障隐患的响应准则，对比各故障特征量在隐患工况和正常工况时的幅值，探讨了各故障特征对不同隐患工况的识别能力，初步验证了隐患特征对不同隐患进行挖掘的有效性，为利用在途数据进行隐患挖掘提供了一个可行途径.

Abstract:

In order to dig the hidden trouble of subway train’s service air braking, the following performance of cylinder volume (CV) actual pressure and CV target pressure is chosen to be the fault feature. According to the change of CV actual pressure of the running train, service air braking process is divided into three typical braking phases: decelerating phase, hold off brake phase and release phase. For each of the three phases with own characteristic a corresponding method is put forward to extract fault feature, including correlation method, mean absolute error method and direct extraction, and there are totally six fault features extracted. At last, the three kinds of typical faults are tested on the brake system reliability test. By means of the criterion of response to the faults, amplitudes of the six fault features in hazard conditions and in normal working condition are compared. And recognition ability of the six fault features are discussed. The effectiveness of the six fault features for excavating the different hidden troubles is established via the comprehensive analysis of the running data and the fault simulation data.