Increase in accuracy forecast for electrical energy consumption of the WEM subjects using fuzzy neural networks

Abstract

This paper covers in depth functioning essence of the balancing electricity market in the Russian Federation in accordance with the current economic stimuli. Significant role of the prediction process for energy resources consumption has been noted, which has a direct influence on a pricing policy within the marked trade area. Analysis of different factors, that impact execution accuracy of the one day in advance model, was conducted. Problem of the weakly defined and unknown parameters, that influence the performance of the balancing electricity market, is suggested to resolve by means of fuzzy neural controller. The paper presents its structure and settings, and possibility to scale it to use not only as an analyzer but as an executive element in the subsystems of electric power systems. High accuracy performance of the fuzzy neural controller was shown - around (0,5-3) %, that is much higher than methods already in use (for instance, based on a coefficient of decline/growth of power consumption).