Optimal interpolating windowed discrete Fourier transform algorithms for harmonic analysis in power systems

An interpolating windowed discrete Fourier transform (DFT) has been used to eliminate the errors caused by the leakage and picket fence effects associated with use of a conventional DFT. An interpolating algorithm eliminates the errors caused by the picket fence effect, and the errors produced by the leakage effect are reduced by windowing the signals. The precision of the harmonic analysis and the implementation of the interpolating algorithm are both affected by the choice of window function. The leakage effect of a conventional DFT and the windowed DFT are analysed. A method to construct optimal window functions and the corresponding interpolation algorithms is presented. Simulation results show that the interpolating windowed DFT algorithms have a high accuracy and are easy to implement and they are therefore useful tools for precise and practical harmonics evaluation in power systems.