基于子波变换和AR模型的轴心轨迹识别

IDENTIFICATION OF THE ORBITS OF SHAFT CENTERLINE BASED ON WAVELET TRANSFORMS AND AR MODEL

应用子波及子波向量，结合自回归时间序列ＡＲ模型，给出了大型旋转机械轴心轨迹的识别方法．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．实验表明本方法具有较高的识别率．

In this paper, based on wavelet and wavelet vectors and combined with AR model of time series, we give an identification algorithm of the orbits of shaft centerline for large rotating machinery. The algorithm can not only contain the global features of the orbits of shaft centerline, but also describe the correlation of x axial and y axial signals to a certain degree. Experiments show that the algorithm possesses higher identifying rate.