自相关函数的多普勒信号频率估计算法

Doppler Signal Frequency Estimation Algorithm Based on Autocorrelation Function

针对激光多普勒测振系统中多普勒信号强度弱、信噪比低的特点,在自相关函数法的基础上,提出了用于频率估计的归一化自相关函数法.通过对自相关函数归一化以提高算法的鲁棒性,同时对频率估计的精度进行评估.采用激光多普勒速度干涉仪测量音叉振动,使用该算法对多普勒信号进行频率估计.实验结果表明:信号频率曲线和音叉速度曲线与音叉的标称频率相符,在多普勒信号的幅值和信噪比变化时,归一化自相关函数检测法具有极强的鲁棒性,用于多普勒频率估计时精度高,对噪声的容忍能力较强.算法的频率估计精度受信号频率变化率影响,信号频率变化率越小,频率估计精度越高.

For the low intensity of shift signal and low signal-to-noise ratio in laser Doppler vibrometer system,a normalized autocorrelation function method for frequency estimation, on the basis of the autocorrelation function, is put forward. Through the normalization of autocorrelation function to improve the robustness of the algorithm, and to evaluate the accuracy of frequency estimation. The laser Doppler velocity interferometer is used to measure the vibration of a tuning fork, and this algorithm is used to estimate the frequency of Doppler signal. The simulation results show that the signal frequency curve calculated by the algorithm and the tuning fork speed curve are consistent with the nominal frequency of the tuning fork. When the amplitude of the Doppler signal and the signal to noise ratio change, normalized autocorrelation function method has strong robustness. It has high precision for Doppler frequency estimation and strong toleration for noise. The frequency estimation accuracy of the algorithm is affected by the change rate of the signal frequency. While change rate of the signal frequency is smaller, the frequency estimation accuracy is higher.