Multifactor Regression Model for Forecasting Production Power Consumption in Electric Grids

Abstract

The article presents the results of the development of a multifactor regression model for forecasting electric power consumption (losses) in production based on the historical data from the key elements of the electric energy balance: electricity consumption and net power supply values, as well as the electric power loss value and level. Apart from the regression model of forecasting, this research provides some corrections to the forecast value that take into account the significant factors affecting the actual power consumption in production, such as the temperature, average electric power loss value and level, and the expected power grid supply dynamics. The developed model was tested in practice and the observed forecasting accuracy made up over 98%, which means that the model can be used by electric grid companies.