时序多相关-经验模式分解方法及其对车辆振动信号的分析

Multi-Correlation of Time Series and Empirical Mode Decomposition and Vibration Signal Analysis of Vehicle

振动信号的特征提取由于受强背景噪声的干扰往往具有很大的困难。在作者提出的时序多相关-经验模式分解方法的基础上,提出了一种相应的改进方法,将其扩展到一般振动信号的特征提取。先对采集到的时间序列作多相关处理,在时序多相关处理时,为了达到克服噪声干扰的目的,取一段采样序列,对其作周期延拓,使得在多相关处理后,噪声仅仅体现在多相关序列的常数项里面。再对得到的多相关数据作经验模式分解,选择满足要求的本征模式函数并作边际谱分析,以达到提取强噪声背景下的特征信号的目的。仿真分析表明了该方法的有效性。最后将它应用到实际某特种车辆振动信号的特征提取中,得到了满意的结果。

The strong background noise always makes great difficulty to the feature extraction of the vibration signal. The multi-correlation of time series and empirical mode decomposition （MCTS-EMD） is extended and used to extract the feature of the vibration signal of the common machine. Firstly, one sampling of the time series is selected, and the periodic extending is the next step, then the multi-correlation process is made for the time series. In order to overcome the interference of the zero-mean noise disposing MCTS, the noise only appears in the constant terms for MCTS. The feature extraction of the vibration signal is obtained in the strong background noise by EMD after MCTS. The simulation analysis proves that the method is effective. Finally, the extended MCTS-EMD is used to extract the feature signal of a special vehicle and it is in agreement with the analysis result.