Harmonic Parameter Online Estimation in Power System Based on Hann Self-Convolving Window and Equidistant Two-Point Interpolated DFT

摘要

A revised interpolated DFT is proposed to estimate the harmonic parameters accurately. Firstly, a 4-order Hann self-convolving window is employed to restrain the interferences caused by spectral leakage and sidelobes. The corresponding error can be restrained at a degree of 10(-15) theoretically. Then, an equidistant two-point interpolated formula is proposed to deal with the picket-fence effect. It avoids the systematic error caused by the polynomial fitting in common interpolation method. For estimating precision, the theoretical relative errors of harmonic parameters would approximate 10(-11). The relative frequency error of the fundamental wave is about 10(-7) in a 50-dB-noise environment. Moreover, in a 20-dB-noise background, the signal reconstructed could reduce the variances to 0.001 of its own, which is feasible and useful in practical harmonic detection.