Neural network for estimation of harmonic components in a power system

Abstract

The paper presents a novel approach to the estimation of harmonics in the power system. The proposed method uses the optimisation neural network theory and its implementation in an analogue form. As is proven in the paper, this formulation, and its practical implementation, leads to the correct solution of the problem in real time. The solution presented is inexpensive and suitable for use with the solid state compensating systems to improve the quality of delivered energy.