结合矢谱与谐波小波包的信息融合方法

Data Fusion Based on the Vector Spectrum and Harmonic Wavelet Packet

旋转机械在运行时产生的信号频率复杂,而在测点上获取的信号是单源信号,只是反映某个方向上的频率成分和幅值大小,难以表述一个截面上的整体信息,在机械故障诊断时容易产生误判。本文将全矢谱技术和谐波小波包方法结合,旨在对一个截面上两个同源非平稳信息进行融合,反映一个截面综合信息。文章最后以仿真信号验证矢谐波小波包信息融合方法的正确性。

Because the running rotating machines have various signal frequency, the signal from single channel only embodies frequency and amplitude value of that direction. It cannot completely represent the information of this section. So mechanical fault diagnosis are likely to have imprecise judgment. In the paper, the vector spectrum is combined with the wavelet packet to fuse the homologous non - stationary signals, representing the information of this section. Emulation signal is used to validate data fusion based on the vector spectrum and harmonic wavelet packet.