针对激光多普勒信号中存在较大噪声干扰的实际情况，为了抑制这些噪声干扰，提高激光多普勒测速仪的测量精度，提出了对激光多普勒信号进行最小均方差(LMS)自适应滤波后作快速傅里叶变换(FFT)，基于混合编程思想对所得到的频谱，先进行频谱细化，再进行频谱校正的信号处理方法，并对理想正弦信号和实测多普勒信号分别进行仿真计算和实验研究。仿真和实验结果表明：LMS自适应滤波技术可以有效抑制激光多普勒测量中的多频率噪声的干扰，此技术能够适应于很宽的信噪比范围，大大提高多普勒信号的信噪比；频谱细化技术可以提高激光多普勒信号的频谱分辨率，频谱校正技术可以准确地校正多普勒频率，使校正后的频率更加接近于真实值；信号处理精度比直接进行FFT提高2~3倍。

A new method of signal processing is proposed in this article in order to eliminate the noise interference and enhance the measurement accuracy of laser Doppler velocimeter (LDV). The least mean square (LMS) adaptive filter technology, the frequency spectrum zooming and frequency spectrum correction technology are applied in the signal processing system. Basic principle of LMS adaptive filter and frequency analysis algorithms were expounded first, then spectrum simulation of ideal sinusoidal signal and experiment test are studied based on these algorithms. The simulation and experiment results indicate that the LMS adaptive filter technology has effectively improved signal to noise ratio and minish the interference of noise; Resolution of frequency spectrum can be improved by frequency spectrum zooming technology and the accuracy of signal processing can be greatly improved by frequency spectrum correction technology. It can obviously increase the frequency spectrum resolution, and improve the measurement accuracy for 2~3 times.