Techniques for identification of harmonics in industrial power systems

Two techniques for accurate harmonic analysis of industrial power systems are presented. The first is based on the fast Fourier transform (FFT) of the sampling sequences of voltage and current waveforms. It performs a decimation of the acquired data, adopting nonrecursive low-pass FIR filters, to achieve frequency zooming in the spectrum. Then, a flat-top windowing is executed to reduce the negative effects of leakage and interference phenomena. The second technique is based on the extended Kalman filter (EKF) theory. It permits online identification of the instantaneous values of the fundamental and harmonic contents. It also gives the r.m.s. harmonic contents and the total harmonic distortion factor. Measurements were conducted on electronic power converters, which are the major sources of harmonic in industrial environment. The results confirmed the high accuracy of estimates.<>