DOI: https://doi.org/10.15407/publishing2017.48.005

INCREASING THE ACCURACY OF SHORT-TERM ELECTRICAL LOAD FORECASTING WITH CONSIDERING TO CHANGES IN THE STRUCTURE OF CONSUMPTION DURING THE YEAR

P. Chernenko, V. Miroshnyk
Institute of Electrodynamics of the National Academy of Sciences of Ukraine,
Peremohy, 56, Kyiv-57, 03680, Ukraine,
е-mail: This e-mail address is being protected from spambots. You need JavaScript enabled to view it

The paper presents an analysis of the influence of annual periodicity on the accuracy and stability of the electrical load short-term forecasting results. Two approaches are considered which take into account the different behavior of the electrical load in the heating season and off-season. For forecasting, we used the multilayer perceptron with scaled exponential linear unit (SELU) function used as a nonlinear transformation in hidden neurons. This function stabilizes mean and variance of layers and accelerates the learning process. In the first approach, the neural network included an additional input neuron that takes values of 1 for days that are part of the heating season and 0 for the off-season days. In this case, the given model fitted on the annual data. In the second approach, two separate neural networks are developed for work in different seasons of the year. Input vector was generated separately for each network. Estimation of the accuracy and stability of the forecasting results was carried out on year data for real electricity supply company. References 8, figures 3, table 1.
Key words: electrical load, annual periodicity, short-term forecasting, artificial neural network.

1. Chernenko P.A. Parameters’ identification, design and multi-level interconnected prediction of electrical loads of power interconnection. Tekhnichna elektrodynamika. Problemy suchasnoi elektrotekhniky. 2010. Part 3. P. 57–64. (Rus)
2. Brodowski S., Bielecki A., Filocha M. A hybrid system for forecasting 24-h power load profile for Polish electric grid. Applied soft computing. 2017. Vol. 58. P. 527–539. DOI: https://doi.org/10.1016/j.asoc.2017.04.053
3. Ceperic E., Ceperic V., Baric A. A strategy for short-term load forecasting by support vector regression machines. IEEE Transactions on power systems. 2013. Vol. 28, N 4. P. 56–64. DOI: https://doi.org/10.1109/TPWRS.2013.2269803
4. Jones E., Oliphant E., Peterson P. et al. SciPy: Open Source Scientific Tools for Python, 2001. http://www.scipy.org/
5. Kingma D.P., Ba J. Adam: A method for stochastic optimization. 3rd International Conference for Learning Representations, San Diego, 2015.
6. Klambaue, G., Unterthiner T., Mayr A., Hochreiter S. Self-Normalizing Neural Networks. In Advances in Neural Information Processing Systems, 2017.
7. Nocedal J., Wright S.J. Numerical Optimization: Second Edition. Springer, New-York, 2006. P. 664.
8. Suganthi L., Samuel A.A. Energy models for demand forecasting. Renewable & sustainable energy reviews. 2012. Vol. 16, N 2. P. 23–40.  DOI: https://doi.org/10.1016/j.rser.2011.08.014