Development of neuronet models to predict voltage deviations in electrical networks

Abstract

The article examines the increase of effectiveness of operation of electrical equipment with the help of neuronet model for hourly prediction of voltage deviation in electrical networks of industrial and residential areas, taking into account power quality. Due to the constant increase in power consumption and increase of its cost, interest to conserve electricity becomes much higher. The upcoming significant increase of power consumption in the cities of Ukraine for the period until 2030, and the increase in urban population confirm the urgency of the problem. Further saturation of flats by electricity receivers of increased power will sharpen the issues of power quality in electrical networks, so there is a certain interest in neuronet models, which have been successfully used in various fields of science and technology. Neural networks present a modern method of simulation, which permit to reproduce extremely complicated dependences. The proposed model indicates the possibility of its usage as a reference one in the systems of voltage control. Reduction of power consumption in electrical units will save electricity and reduce peak power