针对实际激光多普勒速度信号中混有大量噪声信号、难以找到所需多普勒频移的问题, 提出了一种小波阈值消噪方法。根据多普勒信号的特点, 通过选取合适的小波分解层数、小波基函数、阈值函数和阈值估计方法, 实现小波阈值消噪的最优效果。Matlab仿真和实验结果表明, 小波阈值消噪可以快速有效地消除多普勒信号中的噪声信号, 提高信噪比, 得到有用信号, 有助于准确地寻找到所需的多普勒频移。小波阈值消噪方法中的分层阈值消噪效果优于全局阈值消噪。

The actual laser Doppler velocity signal is always mixed with noise, which leads to great difficulty to find required Doppler frequency shift. In view of this problem, a wavelet threshold de-noising method is presented. According to the characteristics of the Doppler signal, the optimal effect of wavelet threshold de-noising is obtained when we choose proper wavelet decomposition layer number, wavelet basis function, threshold function and threshold estimation method. Results of Matlab simulations and experiments show that the wavelet threshold de-noising method can eliminate the noise in Doppler signals quickly and effectively, so that the signal to noise ratio is improved and the useful signal is obtained. In addition, the method is helpful to find required Doppler frequency shift more accurately, and the hierarchical threshold de-noising effect in wavelet threshold de-noising method is better than the global threshold de-noising effect.