转向架齿轮箱动力学建模与故障量化分类方法研究

Research on dynamic modeling and fault quantitative classification method of bogie gearbox

精确的故障诊断技术是运营安全性和维护高效性的重要支撑,基于动力学建模和频域特征分析,提出齿轮故障的波峰计数量化分类方法,以定量的方式实现故障类型的精确区分。首先,建立可以精确区分正常及擦伤、缺角和齿根裂纹等类型故障的齿轮动力学模型,获得齿轮故障的动力学响应特征,然后基于模型仿真信号的频域特征,提出精确区分不同类型故障的波峰计数诊断方法,最后,以转向架高保真故障模拟试验台开展齿轮故障模拟试验,验证动力学模型特征分析的准确性和量化诊断方法的有效性。研究结果表明:在一定的故障严重程度下,裂纹、擦伤、缺角、正常这4种状态的波峰计数依次减小,利用波峰计数方法可准确划分故障类型。本文研究为转向架齿轮故障动力学特性分析和故障诊断提供了重要方法。

Accurate fault diagnosis technology is an important support for operation safety and maintenance efficiency.Therefore,based on Dynamic Modeling and frequency domain feature analysis,an accurate diagnosis method of gear fault by peak counting was proposed,which can accurately distinguish the fault types in a quantitative way.Firstly,a gear dynamic model was established,which can accurately distinguish the states of normal,scratch,tip broken and root crack,and the dynamic response characteristics of gear faults were obtained.Furthermore,based on the frequency domain characteristics of simulation signals,a peak count diagnosis method for accurately distinguishing different types of faults was proposed.Finally,the gear fault experiment was carried out on the high fidelity bogie test-bed,which verified the accuracy of the dynamic model and the effectiveness of the accurate diagnosis method.The results show that at a certain degree of fault severity,the peak counts of cracks,scratches,tip broken,and normal four states decrease sequentially.The peak counting methods can accurately classify the types of faults.The research in this paper provides an important method for fault dynamic characteristic analysis and fault diagnosis of bogie gears.