一种基于改进DFT算法的相位差测量研究

A Research on Phase Difference Measurement Based on Improved DFT Algorithm

相位差是指两路同频率周期信号初相位的差,在工业系统中有广泛应用,传统DFT算法测量工频相位差存在很大的频谱泄露误差。论文提出一种改进DFT相位差算法,使用布莱克曼自卷积窗对信号截短,并利用离散频谱校正方法修正有用频谱处的相位,并计算出相位差。仿真实验证实,在存在多次谐波干扰条件下,改进DFT算法能够克服频率波动和白噪声干扰,算法精度比现有的矩形自卷积窗插值算法提高2个数量级以上,可以实现相位差的高精度测量。

Phase difference refers to the difference of two equal frequency periodic signals＇ initial phase, which have been extensively applied in the industrial systems. The traditional DFT algorithm has great spectrum leakage when measuring the phase difference of industrial frequency signals. The paper provides an improved DFT algorithm, which uses a Blackman self-convolution window to shorten the signal, corrects the phase of useful spectrum with discrete spectrum correction method and calculate the phase difference. Simulation experiments show that the improved DFT algorithm can overcome the frequency fluctuation and white noise interference when adding multiple harmonics, the accuracy precision of improved DFT algorithm has two orders of magnitude above than existing algorithm named rectangle self-convolution window interpolation algorithm, so it can achieve high accuracy for phase difference measurement.