Spectral interpolation for frequency measurement at off-nominal frequencies

Abstract

Frequency is one of the most vital parameters used in power system monitoring, protection, and control. The commonly used frequency domain methods for frequency estimation are based on the compensation of DFT spectrum. Some accurate frequency estimation methods utilize the linearization of local DFT spectrum, such as interpolated DFT, primarily aiming at higher estimation accuracy of local maxima of signal spectrum. The accuracy of these methods is questionable due to the distortion from spectrum leakage and picket fence effects. An algorithm that estimates the frequency of quasi-steady state sinusoidal signals is presented in this paper. The proposed algorithm produces accurate frequency estimation by calculating DTFT parameters using nonlinear interpolation method on DFT spectrum. IIR filter is utilized for signal conditioning. To expedite iteration process, autocorrelation function of the original signal is analyzed. Test results show that the proposed algorithm achieves higher accuracy than the requirements from IEEE standard C37.118.1a-2014.