Abstract
Nowadays, harmonic distortion in power systems is attracting significant attention. Traditional technical tools for harmonic distortion analysis using either fast Fourier transform or discrete Fourier transform are, however, susceptible to the presence of noise in the distorted signals. Harmonic detection by using Fourier transformation also requires input data for more than one cycle of the current waveform and requires time for the analysis in the next coming cycle. In this paper, an alternative method using neural network algorithm has achieved satisfactory results for fast and precise harmonic detection in noisy environments by providing only 1/2 cycle sampled values of distorted waveforms to neural network. Sensitivity considerations are conducted to determine the key factors affecting the performance efficiency of the proposed model to reach the lowest errors of testing patterns

This publication has 17 references indexed in Scilit: