通过对连续小波变换的分析研究 ,提出了一种提取信号在小波尺度上的能量谱的信号分析方法。该方法能有效地对不同磨损状况下的齿轮振动信号进行分析 ,分析结果说明信号在小波尺度上的能量谱与齿轮的磨损程度有密切的关系。求出不同磨损状况下齿轮振动信号的能量谱对尺度的积分值 ,并根据这些值拟合得到的曲线与齿轮磨损过程曲线非常相似 ,这说明可以用连续小波变换的能量谱估计齿轮磨损状况。最后提出了一种连续小波变换的齿轮磨损特征量提取方法 ,用于提取齿轮磨损程度的特征向量 ,特征量间的欧氏距离说明这些特征向量能很好地表征齿轮的磨损状况 Based on the analysis of continuous wavelet transform, a signal analysis method of extracting the energy spectrum of the signal on the wavelet scale is proposed. The method can effectively analyze the gear vibration signal under different wear conditions. The analysis results show that the energy spectrum of the signal on the wavelet scale is closely related to the gear wear degree. The integral of the energy spectrum of the gear vibration signal to the scale under different wear conditions is obtained. The curve fitted by these values ​​is very similar to the gear wear curve, which shows that the gear wear state can be estimated by the energy spectrum of the continuous wavelet transform. At last, a method of gear wear feature extraction based on continuous wavelet transform is proposed, which is used to extract the eigenvectors of the degree of gear wear. The Euclidean distances between the eigenvalues ​​show that these eigenvectors can well characterize the gear wear