Estimation Algorithms of Harmonic Parameters Based on the FFT

The spectral leakage of the fast Fourier transform (FFT) algorithm caused by asynchronous sampling seriously affects the parameter estimation accuracy. The interpolation algorithm based on window function and the phase difference algorithm is presented in this paper to amend the FFT results by calculating the synchronous deviation. The interpolation algorithm is commonly used in harmonic analysis, but it is difficult to apply to some cosine windows with high-order or certain coefficient, because it needs to solve high-order equations to obtain the explicit interpolation formula. The Phase difference algorithm is applicable to cosine window of any order, which uses the phase difference between the discrete Fourier transform (DFT) spectrum peaks of two adjacent sampled data to estimate the synchronous deviation. Simulation results are compared and analyzed to verify the performance of the two algorithms.