Electricity consumption at the education institutions studied is expected to increase which means that a more efficient and relevant model for managing electricity consumption is needed. To achieve this, statistical forecasting technique was applied. An accurate forecast of consumption would be great help for executives and for those concerning to establish a power-saving policy in the universities. The objective of this research was to forecast electricity consumption by using hybrid models (mixing the Box-Jenkins modeling and a support vector regression model). Three Rajabhat Universities: Nakorn Ratchatsima, Ubon Ratchathani and Loei, were included in the modeling. Time series data on the energy consumption on a monthly basis, covering the period from January, 2006 to September, 2017, were available on the website of the Energy Ministry. Data for the period of January, 2006 to December, 2016 were used for our research purposes, and R-language based forecast model that we developed was used for analysis. The most suitable model was selected by using the Mean Absolute Percent Error (MAPE). The most suitable model obtained from the data series from January to September, 2017 was used to measure the accuracy of the forecast produced. Our results indicated that the proposed model was more efficient, with greater accuracy, in forecasting the energy consumption than the conventional model currently used in the universities participating in the study. For these universities, the MAPE was 7.65428 for Nakorn Ratchatsima Rajabhat University, 6.35679 for Ubon Ratchathani and 4.13581 for Loei Rajabhat Universities.
Keywords: Electricity Consumption, Hybrid models, Box-Jenkins Model, Support Vector Regression Model
Junsagoon, S. (2015). Time Series Forecasting : Linear Approach, Non – Linear Approach and Hybrid Models. Eau Heritage Journal Science and Technology, 9(2), 50–63.
Meyer, D., Dimitriadou, E., Hornik, K., Weingessel, A., Leisch, F., Chang, C.C., & Lin, C.C. (2015). Misc Functions of the Department of Statistics, Probability Theory Group. Retrieved from http://cran.r-project.org/web/packages/e1071/index.html,5 December
Ministry of Energy. Government Energy Reduction Program. (2017). Retrieved form http://www.e-report. energy.go.th/data/index.php.
Singchai, P., & Keeratiwintakorn, P. (2014). Electricity Demand Forecast for Thailand Demand Side Management Center. Information Technology Journal, 10(2), 32–42.
Sujjaviriyasup, T. (2017). Hybrid Model of Support Vector Machine and Genetic Algorithm for Forecasting the Annual Peak Electricity Demand of Thailand. Journal of Kmutnb, 27(3), 451–463.
Theeraviriya, C. (2017). A Comparison of the Forecasting Method for Electric Energy Demand in Nakhonphanom Province. Naresuan University Journal: Science and Technology, 25(4), 124–137.
Vapnik, V. (1998). The Nature of Statistical Learning Theory. Retrieved from https://www.springer.com/ gp/book/9780387987804.
Yashlan, Y., & Bican, B. (2017). Empirical mode decomposition based denoising method with support vector regression for time series prediction: A case study for electricity load forecasting. Measurement, 103, 52–61. https://doi.org/10.1016/j.measurement.2017.02.007.
Zhang, G. P. (2003). Time series forecasting using a hybrid ARIMA and neural network model. Neurocomputiong, 50, 159–175. https://doi.org/10.1016/S0925-2312 (01)00702-0
Zhang, F., Deb, C., Lee, S. E., Yang, J., & Shah, K. W. (2016). Time series forecasting for building energy consumption using weighted Support Vector Regression with differential evolution optimization technique. Energy and Building, 126, 94–103. https://doi.org/10.1016/j.enbuild.2016.05.028
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.