在讨论特征分析方法原理的基础上,针对机车走行部故障在线监测过程中存在的信号分析与处理问题,运用整周期等角度采样方法将时域振动信号转换为角域信号,采用FFT变换将角域信号变换为对应的特征频谱,通过谱估计、谱图分析得到机车走行部各零部件的故障特征谱值,再根据该特征谱值识别机车走行部各零部件的故障。然后,根据机车走行部故障诊断的实际需要,设计了一套基于特征分析方法的机车走行部故障在线诊断系统。实验结果表明,该方法能准确、可靠地识别机车走行部故障。
Aiming at the signal processing problems that exist in online fault diagnosis for train running gear, based on discussing the principle of characteristic spectrum analysis, a method of full period and uniform angle sampling is applied, which transfers the vibration signal from time domain into angular domain. The angular domain signal is transferred into corresponding characteristic spectrum using FFT. Through spectrum estimation and analysis, fault characteristic spectrum values of running gear components are acquired to distinguish their faults. An integrated online fault diagnosis system of train running gear based on characteristic spectrum analysis was designed for actual needs. The experiment results indicate that the method can distinguish the faults of train running gear accurately and efficiently.