STM Journals Peer-reviewed research from Consortium e-Learning Network

Trends in Electrical Engineering Review Article

Support Vector Machine Inspired Load Forecasting of a State University in Haryana

  1. Dheeraj Maharishi Dayanand University, Rohtak
  2. Neha Khurana Maharishi Dayanand University, Rohtak
  3. Pradeep Singla Panipat Institute of Engineering & Technology

Abstract

Estimating the possible environmental impact and determining probable capital requirements are made easier with a solid grasp of electricity demand. Beginning in the middle of the 20th century, demand forecasting for electric power networks was studied theoretically. Prior to that, the study of demand forecasting had not developed because of the small scale of power networks. With the use of statistical prediction techniques, plans for the electric power industry have been created. For fuel management, maintenance scheduling, and budget planning, electric utility companies require monthly peak and annual load forecasts. The approach based on regression models, neural networks, and support vector machines for load forecasting—that is, the maximum and minimum load of the Maharshi Dayanand University (MDU) substation in Rohtak, Haryana—is compared in this paper. Many contributing elements have been investigated and tried. The system created to forecast the greatest and minimum electric demand and consumption in the MDU is presented in the study. Comparing the selected system to other methods in the study, it was found that the support vector machine (SVM)-based model forecasted the load most accurately. The study conducted at the MDU in Rohtak, Haryana, focuses on forecasting the maximum and minimum electrical loads at the university’s substation. It compares three machine learning techniques: regression models, artificial neural networks (ANNs), and SVMs. The findings indicate that the SVM-based model outperforms the others in predicting both peak and minimum loads. SVMs are particularly effective in handling nonlinear, high-dimensional datasets, which are common in electrical load forecasting. They are less prone to overfitting compared to neural networks, especially when dealing with limited or noisy data. SVMs also offer robust generalization capabilities, making them suitable for scenarios where accurate predictions are crucial for operational planning and energy management. In contrast, regression models, while straightforward and interpretable, often struggle with capturing complex, nonlinear relationships inherent in load data. ANNs can model nonlinear patterns but require extensive data and careful tuning to avoid overfitting, and they may demand significant computational resources. SVMs balance complexity and performance, offering accurate predictions without the extensive data requirements of ANNs. Accurate load forecasting is vital for the MDU’s substation to ensure efficient energy distribution, maintenance scheduling, and budget planning. Implementing an SVM-based forecasting model can lead to better resource allocation and operational efficiency, ultimately supporting the university’s energy management objectives.

Keywords

References (28)

  1. International Energy Agency. World Energy Outlook 2009. World Energy Outlook. 2009. doi:10.1787/weo-2009-en
  2. Bunn D, Farmer ED. Comparative models for electrical load forecasting. New York, NY: John Wiley and Sons Inc.; 1984 Dec. Available from: https://www.osti.gov/biblio/6256333
  3. Bunn DW. Forecasting loads and prices in competitive power markets. Proceedings of the IEEE. 2000;88(2):163-169. doi:10.1109/5.823996
  4. Douglas AP, Breipohl AM, Lee FN, Adapa R. Risk due to load forecast uncertainty in short term power system planning. IEEE Transactions on Power Systems. 1998;13(4):1493-1499. doi:10.1109/59.736296
  5. Gross G, Galiana FD. Short-term load forecasting. Proceedings of the IEEE. 1987;75(12):1558-1573. doi:10.1109/proc.1987.13927
  6. Box GE, Jenkins GM, Reinsel GC, Ljung GM. Time Series Analysis: Forecasting and Control. New York, NY: John Wiley & Sons; 2015.
  7. Chen JF, Wang WM, Huang CM. Analysis of an adaptive time-series autoregressive moving-average (ARMA) model for short-term load forecasting. Electric Power Systems Research. 1995;34(3):187-196. doi:10.1016/0378-7796(95)00977-1
  8. Vemuri S, Hill D, Balasubramanian R. Load forecasting using stochastic models. In: Proceedings of the 8th PICA Conference; 1973; Minneapolis, MN. Paper No. TPI-B. p. 31–7.
  9. Wang H, Schulz NN. Using AMR data for load estimation for distribution system analysis. Electric Power Systems Research. 2006;76(5):336-342. doi:10.1016/j.epsr.2005.08.003
  10. Islam BU. Comparison of conventional and modern load forecasting techniques based on artificial intelligence and expert systems. Int J Comput Sci Issues. 2011;8:504.
  11. Zheng H, Yuan J, Chen L. Short-term load forecasting using EMD-LSTM neural networks with a XGBoost algorithm for feature importance evaluation. Energies. 2017;10:1168. doi:10.3390/en 10081168.
  12. Hippert HS, Pedreira CE, Souza RC. Neural networks for short-term load forecasting: a review and evaluation. IEEE Transactions on Power Systems. 2001;16(1):44-55. doi:10.1109/59.910780
  13. Ahmad A, Javaid N, Mateen A, Awais M, Khan ZA. Short-Term Load Forecasting in Smart Grids: An Intelligent Modular Approach. Energies. 2019;12(1):164. doi:10.3390/en12010164
  14. Amara F, Agbossou K, Dubé Y, Kelouwani S, Cardenas A, Hosseini SS. A residual load modeling approach for household short-term load forecasting application. Energy and Buildings. 2019;187:132-143. doi:10.1016/j.enbuild.2019.01.009
  15. Hong T, Fan S. Probabilistic electric load forecasting: A tutorial review. International Journal of Forecasting. 2016;32(3):914-938. doi:10.1016/j.ijforecast.2015.11.011
  16. Nti IK, Samuel AA, Michael A. Predicting monthly electricity demand using soft computing. Rep J Econ Finance. 2019;6:1967–73.
  17. Stavast P. Prediction of energy consumption using historical data and Twitter [master’s thesis]. Computing Science; 2014. Available from: https://fse.studenttheses.ub.rug.nl/id/eprint/11665.
  18. Ouedraogo NS. Modeling sustainable long-term electricity supply-demand in Africa. Applied Energy. 2017;190:1047-1067. doi:10.1016/j.apenergy.2016.12.162
  19. Kumar J, Gupta R, Saxena D, Singh AK. Power consumption forecast model using ensemble learning for smart grid. The Journal of Supercomputing. 2023;79(10):11007-11028. doi:10.1007/s11227-023-05096-4
  20. Zhou KL, Yang SL, Shen C. A review of electric load classification in smart grid environment. Renewable and Sustainable Energy Reviews. 2013;24:103-110. doi:10.1016/j.rser.2013.03.023
  21. Gonzalez-Briones A, Hernandez G, Corchado JM, Omatu S, Mohamad MS. Machine Learning Models for Electricity Consumption Forecasting: A Review. 2019 2nd International Conference on Computer Applications & Information Security (ICCAIS). 2019:1-6. doi:10.1109/cais.2019.8769508
  22. Rajnish, Saini MK, Saroha S. Solar Power Forecasting using Machine Learning Approaches : A Review. 2023 9th IEEE India International Conference on Power Electronics (IICPE). 2023:1-6. doi:10.1109/iicpe60303.2023.10474992
  23. Mele E. A Review of Machine Learning Algorithms Used for Load Forecasting at Microgrid Level. Proceedings of the International Scientific Conference - Sinteza 2019. 2019:452-458. doi:10.15308/sinteza-2019-452-458
  24. Shah RB. A technological literature review on load forecasting in power system using artificial intelligence. Paripex – Indian J Res. 2019;8:14–5.
  25. Kuster C, Rezgui Y, Mourshed M. Electrical load forecasting models: A critical systematic review. Sustainable Cities and Society. 2017;35:257-270. doi:10.1016/j.scs.2017.08.009
  26. Gulati P, Kumar A, Bhardwaj R. Impact of Covid19 on electricity load in Haryana (India). International Journal of Energy Research. 2020;45(2):3397-3409. doi:10.1002/er.6008
  27. Jacob M, Neves C, Vukadinović Greetham D. Short Term Load Forecasting. Mathematics of Planet Earth. 2019:15-37. doi:10.1007/978-3-030-28669-9_2
  28. Cortes C, Vapnik V. Support-vector networks. Mach Learn. 1995;20:273–97.
Support