A Short-Term Load Forecasting Algorithm Using Support Vector Regression & Artificial Neural Network Method (SVR-ANN)

• Published in: 2020 11th IEEE Control and System Graduate Research Colloquium (ICSGRC) pp. 138 - 143
• Main Authors: Abad, L.A, Sarabia, S.M., Yuzon, J.M., Pacis, M.C.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.08.2020
• Subjects: Artificial Intelligence; Artificial Neural Network; load forecasting; regression; Support Vector Regression
• DOI: 10.1109/ICSGRC49013.2020.9232630

Load Forecasting is the science of predicting the most economical amount of electrical power to be supplied by electrical utility companies. Energy conservation is in paired with this study which is needed for our unending need of power supply given limited source of energy. An effective way on how to do it is by using an Artificial Intelligence algorithm and as per chosen, Support Vector Regression and Artificial Neural Network are the machine learning algorithm which is a good combination for forecasting, classifying, and regression. Results were verified through statistical tools: 1) Mean Absolute Error (MAE); 2) Mean Absolute Percentage Error (MAPE). Using SVR-ANN algorithm, an averaged absolute error of 175.6893MW which corresponds to 2.47% was obtained from the analysis. The promising results of this study could be used as an alternative predicting tool for power system operators.