Enhanced Frequency and Harmonic Estimation for Electric Arc Furnace Measurements: A Least Squares Approach to Mitigate Spectral Leakage

In this paper, a Least Squares (LS) based harmonic and frequency estimation method is proposed to analyze power signals with rapidly changing characteristics in time such as Electric Arc Furnaces (EAFs) as an alternative to the classical Discrete Fourier Transform (DFT)-based methods. Although DFT based harmonic analysis methods provide fast and accurate solutions for stationary signals, solutions become less accurate under frequency deviations especially when the signal is sampled at low rates. This paper presents an approach which can accurately calculate the harmonic components of the signals, when the fundamental periods cannot be expressed as integer numbers due to the low sampling rate. One of the outstanding parts of the proposed approach is that, it is possible to increase the performance of harmonic and interharmonic estimation by increasing the performance of frequency estimation with LS technique in cases of rapidly time varying signals such as EAF currents and voltages. The reliability of the proposed approach is proven with synthetically and realistically generated EAF current signals and the results are also evaluated by running the method on field measurements collected at transformer substations supplying EAF plants both for harmonics and interharmonics estimation. Compatibility of the proposed method with parallel processing is also investigated to prove its usability in real-time applications.