利用HHT和重采样技术分析瞬态机械信号

Analysis of Non-stationary Mechanical Signal Using Hilbert-Huang Transform and Data Resampling Technique

介绍了HHT理论及数据重采样技术的实现方法。分析了利用HHT理论和瞬态信号进行机械故障诊断时常用故障特征提取方法的不足。针对瞬态信号中所包含的故障特征是随时间的变化而变化的特点,研究了将数据重采样技术与HHT理论相结合的瞬态信号分析方法。以齿轮箱齿面磨损故障为例进行了算例分析,通过对照分析原始信号与重抽样信号的分析结果,说明了该方法的有效性。

We present the method for realizing Hilbert-Huang Transform （HHT） theory and data resampling technique and analyze the shortcomings of the conventional fault feature extraction methods for diagnosing mechanical faults using the HHT theory and non-stationary signals. Because the fault features contained in the non-stationary signals are time-varying, we study a non-stationary signal analysis method that combines the data resampling technique with the HHT theory. The diagnosis of the tooth wear fault in a gear box is taken as an example, and contrasfive analysis of the fault＇s original signal and its resampled signal proves the effectiveness of the method.