A Long-Term Forecasting Model of Electricity Consumption Volume on the Example of Ups of the Ural with the Help of Harmonic Analysis of a Time Series

In this article, the model of forecasting of electricity consumption volume is analyzed on the basis of harmonic analysis of a time series. Previously, we establish the presence of a trend component described by a second-order polynomial. Based on the construction of the autocorrelation function, the periodicity of the time series of electricity consumption volume is proved, with periods equal to 1 year and 1 week, due to decrease in electricity consumption in summer because of increase in daylight hours and decrease in production on weekends. The model is tested on actual hourly data of United Energy System of the wholesale market for electricity and power in Russia. The model is tested for adequacy using Fisher's criterion and the coefficient of determination. The introduction of two harmonic components (annual and weekly ones) instead of the generally accepted one reduces the approximation error for the current model from 2,25. The proposed scientific tools are recommended in the entity's operating activity of electric power to forecast the major energy market parameters in order to reduce the fines by improving the accuracy of forecasts.

Keywords

forecasting models; main parameters; energy market; harmonic analysis.

References

1. Mokhov V.G., Demyanenko T.S. Modelling of the Time Series Digressions by the Example of the UPS of the Ural. Bulletin of the South Ural State University. Series: Mathematical Modelling. Programming and Computer Software, 2015, vol. 8, no. 4, pp. 127-130. DOI: 10.14529/mmp150412
2. Mokhov V.G., Demyanenko T.S. The Construction of the Trend Component of the Additive Model of Long-Term Forecasting of the Wholesale Market of Electric Energy and Power of Russia Using the Example of the Ural Energy System. Bulletin of the South Ural State University. Series: Economics and Management, 2018, vol. 12, no. 2, pp. 80-87.
3. Singh S. Pattern Modelling in Time-Series Forecasting. Cybernetics and Systems-An International Journal, 2000, vol. 31, no. 1, pp. 49-65.
4. Nogales F.J., Conejo A.J. Electricity Price Forecasting Through 138 Transferfunction Models. Journal of the Operational Research Society, 2006, vol. 57, no. 4, pp. 350-356.
5. Conejo A.J. Day-Ahead Electricity Price Forecasting Using the Wavelet Transform and ARIMA Models. IEEE Transaction on Power Systems, 2005, vol. 20, no. 2, pp. 1035-1042.