MEASUREMENT OF HARMONICS WITH ADALINE USING HALF-CYCLE DATA

Harmonics are undesirable greater frequencies that are overlaid on the fundamental sinusoidal waveform that resulting in creating a distorted waveform. This distortion forms a non-sinusoidal waveform composed of a set of harmonic frequencies, which is harmful to electrical power supply lines. This paper proposes an ADALINE-based algorithm simulated with MATLAB using half-cycle samples. Widrow and Hoff developed ADALINE, a single-layer neural network. It makes use of the delta rule for training to reduce mean square error. It is mainly used because of its advanced and sophisticated machine learning algorithm. ADALINE is used for the estimation of harmonic components. It has higher functionality to respond without delay due to its learning ability. In earlier papers, one full cycle waveform has been used for harmonic measurement. In the proposed system, only half cycle waveform is used for the estimation of harmonics. This algorithm is simulated using MATLAB.