基于轴箱加速度时频域特征的车轮多边形故障识别

Wheel Polygon Fault Diagnosis Based on the Time and Frequency Domain Characteristics of Axle Box Acceleration

机车车轮的多边形化会在轮轨接触的位置引起异常振动,通过轴箱振动加速度的时频特征可以检测出这种异常振动,从而实现对车轮多边形的故障识别。本文首先构建了考虑轮对柔性的刚柔耦合动力学模型,拟合了不同主导阶次的随机多边形车轮多边形样本,样本径跳值均为0.2 mm。其次在频域内对不同阶次的多边形的频谱进行了分析,在时域内对轴箱加速度的均方根值和时域包络谱的峭度值进行了统计,得出了车轮多边形激励下的轴箱加速度时域特征与频域特征。最后在时频与频域特征的基础上提出了一种车轮多边形故障识别流程。经实测轴箱加速度数据验证,该故障识别流程可以有效检测出车轮多边形的阶次,并能有效识别出车轮多边形化的发展程度。

The abnormal vibration at the contact position of the wheel and rail caused by the polygonalization of locomotive wheels can be detected through the time and frequency domain characteristics of the of the axle box acceleration,which enables the fault diagnosis of the wheel polygon.First,a rigid-flexible coupling dynamic model considering the flexibility of the wheelset is constructed,and the polygonized wheelsets samples at the wheel diameter runout value of 0.2 mm of different orders are fitted.Then,the frequency spectrum of different orders is analyzed,and the root mean square value of the axle box acceleration and the kurtosis value of the time domain envelope spectrum are counted,which derives the time-domain and frequency-domain characteristics of axle box acceleration under different excitations.Finally,on the basis of the time-frequency and frequency-domain characteristics,a wheel polygon fault diagnosis process is proposed.The measured axle box acceleration data proves that this fault diagnosis process can effectively detect the order of wheel polygons and identify the degree of the development of wheel polygonization.