The Electric Parameter Measurement Based On Discrete Fourier Transform

The measurement of electric parameters in power systems is the fundamental of stable operation. However, the nonlinear loads result in harmonic pollution, and new techniques, such as smart grids, need the measurement of electric parameters with new approaches. The accuracy, real-time behavior and applicability of the measurement must be ensured, which has become a research topic.With years of development, the measuring methods based on signal analysis can be divided into two sorts, parametric methods and the nonparametric ones. Among the nonparametric methods, the methods based on discrete Fourier transform have high efficiency, which are widely used, especially in some occasions that meet real-time demands.When the electric parameters are measured, due to the non-synchronous sampling and spectrum leakage caused by limited length of signal, certain errors happen via DFT. To overcome the problems, the accuracy can be improved by interpolation.The measuring method based on discrete Fourier transform can be improved by interpolation both in time domain and in frequency domain. In the time domain, rebuilding the sampling signals by interpolation before DFT can reduce the errors caused by non-synchronous sampling. In the frequency domain, calculating the specific value or the phase difference among spectral lines can correct the electric parameters. Generally, the first step is to estimate the frequency, then the amplitude and phase can be corrected according to the practical frequency and the corresponding window functions.On the basis of the above methods, there can be further improvements. To improve the interpolation method in the time domain, the Hermite interpolation functions which are smooth and continuous is derived, and the truncated error can be reduced by Hermite interpolation algorithm.To improve the interpolation method in the frequency domain, a spectral correction method via phase difference with Hanning window is proposed. After one polynomial transformation on the results of DFT, the attenuation rate of side-lobes can be larger and the accuracy of measurement can be improved. The correction formulas of the improved phase difference correction method are obtained in this paper.The feasibility of the these methods is verified via simulating different signal models at different frequencies. The results show that these above methods have better accuracy, applicability and practicability.