Real-time frequency and harmonic evaluation using artificial neural networks

With increasing harmonic pollution in the power system, real-time monitorin g and analysis of harmonic variations have become important. Because of lim itations associated with conventional algorithms, particularly under supply -frequency drift and transient situations, a new approach based on non-line ar least-squares parameter estimation has been proposed as an alternative s olution for high-accuracy evaluation. However, the computational demand of the algorithm is very high and it is more appropriate to use Hopfield type feedback neural networks for real-time harmonic evaluation. The proposed ne ural network implementation determines simultaneously the supply-frequency variation, the fundamental-amplitude/phase variation as well as the harmoni cs-amplitude/phase variation. The distinctive feature is that the supply-fr equency variation is handled separately from the amplitude/phase variations , thus ensuring high computational speed and high convergence rate. Example s by computer simulation are used to demonstrate the effectiveness of the i mplementation. A set of data taken on site was used as a real application o f the system.