基于小波向量和AR模型的轴心轨迹识别

Identification the Orbits of Shaft Centerline Based on Wavlet Vectors and AR-Model

本文对在不同尺度下小波变换的相关性进行了探讨。基于小波向量和自回归ＡＲ模型给出了大型旋转机械轴心轨迹的一种识别方法。该方法不仅能够反映轨迹的总体特征，而且还一定程度地描述了ｘ轴信号与ｙ轴信号的相关性．实验表明此方法是有效的。

In the paper, the correlation of wavelet transforms at different scale is studied. By wavelet-vectors and AR-model of 2-dimension autoregressive time series, we give an identification algorithm of the orbits of shaft centerline. The algorithm can not only contain the global features of the orbits of shaft centerline, but also generally describe the correlation of x-axial and y--axial sigh.