Artificial neural networks applied to single-phase load harmonic characterization

In this paper, an alternative method based on artificial neural networks is presented for the determination of load current harmonic components in a single-phase electric power system. The first six harmonic components are calculated from current waveforms of an AC controller and a single-phase diode bridge rectifier, which are the most common loads in industrial, commercial and residential applications. The influences of such nonlinear loads in power quality issues are also characterized. The proposed method is compared with truncated FFT. Simulation and experimental results are presented in order to validate the proposed approach.