基于EMD关联维的齿轮箱齿轮状态振动辨识

论述了经验模式分解(EMD)的理论和特性，以及关联维数及其特性，并根据实际齿轮箱振动信号的非平稳特征和简单监测征兆的需要，引入分别适于非线性非平稳信号分析和表征系统观测序列非线性、不规则程度的经验模式分解和关联维数，提出用EMD关联维来辨识齿轮箱振动信号，结果表明：当齿轮处于不同状态时，与齿轮故障密切相关的内在模式函数的关联维数明显不同，EMD关联维方法可以作为齿轮故障的特征提取工具。

Vibration Identification of the Gear Conditions for Gearbox Based on EMD Correlation Dimension

The theory of EMD (Empirical Mode Decomposition), as well as the correlation dimension (CD) and its characteristics, was firstly discussed in this paper. Then, according to the non-stationary feature and simple symptom monitoring of the vibration signals of a practical gearbox, the methods of EMD and CD were pre-sented. These two methods are respectively applied to the analysis of nonlinear and non-stationary signals as well as the measurement of the nonlinearity and non-regularity of the observed time series for a system. An EMD & CD method was finally presented for identifying the vibration signals of a gearbox. The results indicate that, in various gear conditions, the correlation dimensions of the intrinsic mode function related to the gear fault mode are obviously different with each other. This means that the EMD & CD method can be used to extract the fault characteristics of a gear.