Short-term forecasting of prices for the Russian wholesale electricity market based on neural networks

Abstract

The article considers the possibility of using neural networks for the short-term forecasting of electricity prices in the day-ahead market (DAM) based on factors strictly determined for the forecast period. A set of six factors has been determined, which allows an hourly forecast of the DAM price to be constructed for a month in each of the four seasons with a high accuracy. The proposed model shows low average errors in forecasting the price for each hour of the month and in turn allows possible significant price deviations to be anticipated.