一种基于离散傅里叶变换的频率测量算法

An Algorithm for Frequency Measurement Based on DFT

对一种基于离散傅里叶变换(DFT)的频率测量算法进行了分析,该算法认为信号的频率偏移量在邻近的2个周期保持不变,因此在信号频率变化较快时频率计算误差较大。针对此问题,提出了一种改进算法。在计算频率偏移量时,除原方法外,还考虑了频率变化率带来的相邻周期相角差的影响。给出了2种方法以获得该算法所需要的信号频率变化率:一种是利用前次计算所得到的频率变化率,另一种是利用相对固定的频率变化率。仿真计算结果表明,改进后的算法具有较好的频率及频率变化率测量精度。

An algorithm for frequency measurement based on discrete Fourier transform （DFT） is analyzed, in which the difference between frequency deviations of two adjacent cycles of the signal is not considered. So the frequency measurement error is rather big when the signal frequency variation is rapid. To solve this problem, an improved algorithm is proposed, in which the influence of difference between frequency deviations of two adjacent cycles of the signal on frequency deviation measurement is considered using the original method. The improved algorithm takes the rate of frequency variation of the signal as a parameter and two methods to identify the parameter are given. One consists in using the rate of frequency variation obtained in the preceding calculation, and the other in using the relatively fixed rate of frequency variation. The results of numerical simulation show that the improved algorithm can give fairly good frequency and accuracy in measuring the rate of frequency variation.