基于Zoom-FFT的改进Rife正弦波频率估计算法
Improved Rife Algorithm Based on Zoom-FFT for Frequency Estimation of Sinusoid Wave

Rife算法是正弦波频率估计的一种经典算法，但其根本缺陷在于低信噪比且被估计频率接近量化频率点时估计性能差。本文通过分析Zoom-FFT的基本原理，验证了其具有可控的局部频谱放大功能，进而提出了一种改进的Rife频率估计算法。通过对信号进行Zoom-FFT处理实现以被估计频率为中心的较窄频段频谱的大幅度细化和放大，然后利用Rife算法进行精确频率估计。仿真结果表明，该算法具有高于传统Rife及其改进算法的估计精度和抗噪声性能，且对真实频率与量化频率点的位置关系不敏感，但计算复杂度有一定增加。
Rife algorithm is a classical algorithm for frequency estimation of sine wave. Its drawback is that the estimation performance is poor when the true frequency is close to the frequency quantization points or the SNR is low. After analyzing the theory of Zoom-FFT, its amplification function for spectrum is verified. Furthermore, an improved frequency estimation algorithm based on Zoom-FFT is proposed. After the Zoom-FFT process, spectrum of a narrowband centered on true frequency is amplified, and then Rife algorithm is used to finish the accurate estimation of signal frequency. Simulation results indicate that the improved Rife algorithm has higher estimation accuracy and better anti-noise performance than Rife algorithm and some other improved algorithms. Moreover, it is insensitive to the position of the true frequency in the spectrum. However, to a certain extent, the computation load is improved.