A novel approach of harmonic analysis in power distribution networks using artificial intelligence

This research presents a new approach to analyze harmonics in electrical power distribution network through statistical estimation technique of neural networks. The typical power system in current times is encountered by numerous power quality (PQ) problems. This is due to the increased usage of electronic circuitry that involves continuous power switching. The advanced switching and non linear nature of sophisticated electronic equipment deteriorates the quality of supply power thus injecting harmonics into the power distribution network. The power system which is not designed to accept the behavior of the non linear loads thus faces the problems of voltage unbalance, sag, swell, momentary or temporary interruptions and ultimately the total shut down or the black out of the system. The problem of interruption in continuous electric power supply to consumers has affected the safety, reliability and economic efficiency of the power distribution network. In this research artificial neural networks have been used for efficiently predicting the harmonics of a power distribution network. The research highlights the importance of focusing on various power quality parameters specially prevention of harmonics in power distribution networks to achieve sustainable availability of quality supply of electrical power by power utilities for its customers. The outcomes of this research were compared and tested with the field results of a power utility in Victoria, Australia.