Wind Farm Harmonic Analysis Based on Full-Phase Hybrid Window Six-Spectral Line Interpolation

In order to solve the spectrum leakage and fence effect problems caused by asynchronous sampling of signals when using traditional Fourier transform (FFT) for power harmonic signal analysis, this paper proposes a wind farm harmonic error analysis method based on full-phase hybrid window six-spectral line interpolation. First, in order to solve the poor main-sidelobe performance of a single window function in harmonic signal processing, a new hybrid window function is constructed by hybrid convolution of Blackman window and 4-term 5th-order Nuttall window; secondly, in order to overcome the limitations of Fourier transform in phase error detection, the full-phase Fourier algorithm is combined to improve the accuracy of phase error detection; finally, by assigning different weights to the peak spectral lines, the accuracy of peak spectral line interpolation is improved, and through MATLAB simulation, the amplitude, phase and frequency of wind farm harmonics are analyzed. The simulation results show that when no white noise is added, the detection errors of amplitude, phase and frequency parameters are up to <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$10^{-11}, 10^{-12}$</tex> and <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$10^{-15}$</tex> respectively; after adding noise, it is found that the detection error is lower in the high noise range, and the error detection of the three parameters is up to 10-6; from the calculation time, the algorithm in this paper only takes about 0.04 seconds, and the detection efficiency is improved by about <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$10 \%$</tex> compared with the calculation time of other algorithms. The experimental results show that the algorithm in this paper has higher detection accuracy and less detection time, and shows excellent performance when processing signals containing white noise.