Improved High Precision Interpolating FFT for Harmonic Analysis

HUANG Fang-neng,HUANG Cheng-jun,CHEN Chen,JIANG Xiu-chen
DOI: https://doi.org/10.3969/j.issn.1006-6047.2007.09.009
2007-01-01
Abstract:Errors in IFFT(Interpolating Fast Fourier Transform) are analyzed and the signal phase is demonstrated as the main cause.It is pointed out that the algorithm applied to complex signal could not be applied to real signal.A high precision IFFT algorithm is thus presented,which corrects the spectrum of FFT to improve its precision based on FFT.Using the useful information of FFT and considering the spectrum difference between complex signal and real signal,it adopts a series of math transforms to overcome the influence of signal phase.As a result,the precision of harmonic analysis is improved without obvious increase of calculating time.Since the complex iteration and special data sampling are not necessary,the requirements for hardware are not strict in implementation.The precise magnitude,phase and frequency of harmonic components in real signal are achieved by analysis.An instance verifies its correctness and higher precision.
What problem does this paper attempt to address?