Load forecasting using Two-level Heterogeneous Ensemble Method for Smart Metered Distribution System

Document Type : Research Article

Authors

Department of Electrical Engineering, NIT Patna, Patna, Bihar, 800005, India

Abstract

A heterogeneous ensemble method for load forecasting (short-term and mid-term) are proposed here. The proposed approach comprises of a two-level hierarchy of machine learning based methods and classical methods to form the ensemble forecaster, where output of the first-stage forecasters are used as input in the second stage. Artificial Neural Network and Support Vector Regression methods are incorporated in the proposed approach as ML forecasters, whereas Holt’s exponential smoothening and multiple linear regression techniques are included as classical forecasters. The proposed two-level ensemble approach forecasts realistic smart metered data more accurately and efficiently for multiple short-term and mid-term load forecasting scenarios with improved accuracy compared to any individual single stage forecasting methods. The prediction accuracy is shown to improve manifolds for the tested practical system. The proposed model also shows improvements compared to existing aensemble-based model.

Keywords

References

References: 
1.Zhang, P., Wu, X., and Wang, X. “Short–term loadforecasting based on big data technologies,” CSEEJournal of Power and Energy Systems, 1(3), pp. 59-67(2015). DOI: 10.17775/CSEEJPES.2015.00036.
2.Wang, Y., Chen, Q., Hong, T., et al. “Review of smartmeter data analytics: applications, methodologies, andchallenges”, IEEE Trans. Smart Grid, 10(3), pp. 3125-3148 (2018). DOI: 10.1109/TSG.2018.2818167.
3.Tian, C., Ma, J., Zhang, C., et al. “A deep neuralnetwork for short- term load forecast based on LSTMand Convolution neural network”, Energies, 11, 3493(2018). DOI: 10.3390/en11123493.
4.Singh, N., Mohanty, S.R., and Shukla, R.D., “Shortterm electricity price forecast based on environmentally adapted generalized neuron”, Energy, 125, pp. 127–39(2017). DOI: 10.1016/j.energy.2017.02.094.
5.Kazemzadeh, M.R., Amjadian, A., and Amraee, T.“Long term electric peak load forecasting of Azarbaijan regional electricity grid”, Iranian Conference onElectrical Engineering (ICEE) (2020). DOI: 10.1016/j.jup.2019.04.001.
6.Hernandez, L., Baladron, C., AguiarCarro, J.M, et al.“A survey on electric power demand forecasting:Future trends in smart grids, microgrids, and smartbuildings”, IEEE Commun. Surv. Tutorial, 16, pp.1460-1495 (2014). DOI: 10.1109/SURV.2014.032014.00094.
7.Moghram, I. and Rahman, S. ``Analysis and evaluationof five short-term load forecasting techniques", IEEETrans. Power Syst., 4(4), pp. 1484-1491 (1989). DOI: 10.1109/59.41700.
8.Alex, D. and Timothy, C. “A regression-based approach to short term system load forecasting”, IEEE Trans. onPower Syst. 5(4), pp. 1535–1550 (1990). DOI: 10.1109/59.99410.
9.Hagan, M.T. and Behr, S.M. “The time series approachto short term load forecasting”, IEEE Trans. of PowerSyst. PWRS-2(3) (1987). DOI: 10.1109/TPWRS.1987.4335210.
S. Rai and M. De /Scientia Iranica (2025) 32(1): 641010.Rahman, S. and Drezga, I. “Identification of a standardfor comparing short-term load forecasting techniques”,Electric Power Systems Research, 25(3), pp. 149-158(1992). DOI: 10.1016/0378-7796(92)90013-Q.
11.Li, Y., Han, D., and Yan, Z. “Long-term system loadforecasting based on data-driven linear clusteringmethod”, J. Mod. Power Syst. Clean Energy, 6, pp.306–316 (2018). DOI: 10.1007/s40565-017-0288-x.
12.Ji, P.R., Xiong, D., Wang, P., et al. “A study onexponential smoothing model for load forecasting”, InProceedings of 2012 Power and Energy EngineeringConference, Shanghai, pp. 1–4 (2012). DOI: 10.1109/APPEEC.2012.6307555.
13.Gob, R., Lurz, K., and Pievatolo, A. “Electrical loadforecasting by exponential smoothing with covariates”,Applied Stochastic Models in Business and Industry,29(6), pp. 629–645 (2013). DOI: 10.1002/asmb.2008.
14.Filho, K.N., Lotufo, A.D.P., and Minussi, C.R.“Multinodal load forecasting using a general regressionneural network”, IEEE Trans. on Power Delivery,26(4), pp. 2862-2869 (2011). DOI: 10.1109/TPWRD.2011.2166566.
15.Zongying, L., Loo, C.K., and Pasupa, K. “A novelerror-output recurrent two-layer extreme learningmachine for multi-step time series prediction”,Sustainable Cities and Society, 66 (2021). DOI:10.1016/j.scs.2020.102613.
16.Zhang, X., Wang, J., and Zhang, K. “Short-termelectric load forecasting based on singular spectrumanalysis and support vector machine optimized byCuckoo search algorithm”, Electric Power System andResearch, 146, pp. 270-285 (2017). DOI: 10.1016/j.epsr.2017.01.035.
17.Ceperic, E., Ceperic, V., Member, S., et al. “A strategyfor short-term load forecasting by support vectorregression machines”, IEEE Trans. Power System,28(4), pp. 4356–4364 (2013). DOI: 10.1109/TPWRS.2013.2269803.
18.Hamed H.H. Aly, ``A proposed intelligent short-termload forecasting hybrid models of ANN, WNN and KFbased on clustering techniques for smart grid’’, ElectricPower Systems Research, 182 (2020). DOI: 10.1016/j.epsr.2019.106191.
19.Li, S., Wang, P., and Goel, L. “Short-term loadforecasting by wavelet transform and evolutionaryextreme learning machine”, Electric Power SystemsResearch. 122, pp. 96–103 (2015). DOI: 10.1016/j.epsr.2015.01.002.
20.Hendawia, M.E. and Wanga, Z. “An ensemble methodof full wavelet packet transforms and neural networkfor short term electrical load forecasting”, ElectricPower Systems Research, 182, pp. 1-13 (2020). DOI: 10.1016/j.epsr.2020.106265.
21.Polikar, R. “Ensemble based systems in decisionmaking”, IEEE Cicuits and Systems Magazine, 6(3),pp. 21-45 (2006). DOI: 10.1109/MCAS.2006.1688199.
22.Khwajaa, A.S., Anpalagana, A., Naeemb, M., et.al.“Joint bagged-boosted artificial neural networks: Usingensemble ML to improve short-term electricity loadforecasting”, Electric Power Systems Research. 179(2020). DOI: 10.1016/j.epsr.2019.106080.
23.Nazar, M.S., Fard, A.E., Heidari, A., et al. “Hybridmodel using three-stage algorithm for simultaneousload and price forecasting”, Electric Power SystemsResearch, 165, pp. 214-228 (2018). DOI: 10.1016/j.epsr.2018.09.004.
24.Laouafi, A. Mordjaoui, M., Haddad, S., et al. “Onlineelectricity demand forecasting based on an effectiveforecast combination methodology’’, Electric PowerSystems Research, 148, pp. 35–47 (2017). DOI:10.1016/j.epsr.2017.03.016.
25.Palaninathan, A.C., Qiu, X., and Suganthan, P.N.``Heterogeneous ensemble for power load demandforecasting”, IEEE Region 10 Conf (TENCON),Singapore, pp. 2040-2045 (2016). DOI: 10.1109/EPE.2016.7521771.
26.Dudek, G. “Heterogeneous ensembles for short-termelectricity demand forecasting”, 17th InternationalScientific Conference on Electric Power Engineering,Prague, pp. 1-6 (2016). DOI: 10.1109/EPE.2016.7521771.
27.Wang, L., Mao, S., Wilamowski, B.M., et al.``Ensemble Learning for Load Forecasting”, in IEEETransactions on Green Communications andNetworking, 4(2), pp. 616-628 (2020). DOI: 10.1109/TGCN.2020.2987304.
28.Lee, J. and Cho, Y. “National-scale electricity peak load forecasting: Traditional, machine learning, or hybridmodel?”, Energy, 239, 122366 (2021). DOI:10.1016/j.energy.2021.122366.
29.Rai, S. and De, M. “Analysis of classical and machinelearning based short-term and mid-term loadforecasting for smart grid”, International Journal ofSustainable Energy, 40(9), pp. 821-839 (2021). DOI: 10.1080/14786451.2021.1873339.
Scientia Iranica
Volume 32, Issue 1
Transactions on Computer Science & Engineering and Electrical Engineering
January and February 2025 Article ID:6410

Files

History

  • Receive Date: 17 January 2022
  • Revise Date: 16 July 2022
  • Accept Date: 30 January 2023

Share

How to cite

Statistics