Spectrum Interpolated Correction Method And Its Application

Along with the electric power and electronic installation's widespread use in electrical network, the overtone pollution condition is serious day by day. The measurement and analysis of harmonics is the precondition and base of harmonics suppression. Discrete Fourier transform (DFT) is one of the most popular methods applied in computer-based measurement of harmonics signals. However, the non-synchronous sampling, finiteization, and discretization of time-domain and frequency-domain during the spectrum analysis of electric signals cause measuring errors, which makes it difficult to satisfy the request of high detection precision. Therefore, it has important theoretical significance and practical value to engage the research and supplement the remove of harmonic signal spectrum leakage.This paper firstly analyzes the reason of frequency leakage under non-coherent sampling process, concluding that is caused by the phase variations between the original and truncated signals. Traditional solutions are do corrections on spectrum directly, according to the change phase estimated by the window and interpolation algorithm. This process made great improvement in electric power harmonic spectrum resolution; however, it remains large spectrum estimation error on weak high-order harmonic analysis. The recent FXT method based the phase estimate on compute the phase of every spectral line increase the possibility of phase reversal. The algorithm proposed in this paper firstly gets the phase change by the window interpolation method, and then do phase compensation in time domain instead of the frequency domain. The analysis concludes: the truncated data spectral leakages on high order harmonics are compensated when the corrected data sequence is transformed into the frequency domain by the DFT. Based on the analytical result, the paper extend the target signal to a more wide-use situation: how to do effective harmonic research in actual electric power system when the periodical signal contains both Harmonics and Interharmonics. After building a difference filter, the two different kinds of signals will be departed. For the Harmonics part, the time domain spectral compensation method can be applied. For the Interharmonics part, after the phase correction, a new method named group weighted method was proposed to do the non-integer part of harmonic analysis. Both the simulation and experiment results prove the effectiveness and robustness of proposed approaches. Finally, the algorithm is used on the application of ADC dynamic parameters test. The test method of ADC dynamic performances under non-coherent sampling process was interpreted in this paper. Both simulation and experiment results verify the practicability and robust of the proposed test algorithm.