Very short-term load forecasting based on a pattern ratio in an office building

Author(s):Ah-Yun Yoon, Hyeon-Jin Moon, Seung-Il Moon
Department of Electrical and Computer Engineering, Seoul National Univ. #013, 1 Gwanangno, Gwanak-gu, Seoul, 08826, South Korea
International Journal of Smart Grid and Clean Energy, vol. 5, no. 2, April 2016: pp. 94-99
ISSN: 2315-4462 (Print)
ISSN: 2373-3594 (Online)
Digital Object Identifier: 10.12720/sgce.5.2.94-99

Abstract:The pattern of electric demand needs to be analyzed to obtain a simple and precise Very Short-Term Load Forecast (VSTLF) for an office building because the electric demand of a small power system such as a building is difficult to express as a function. In order to develop an improved VSTLF, data from LG Electronics was analyzed. The proposed method is compared to the conventional method using a correlation between electric demand and temperature. The test results show that the proposed method based on a pattern ratio is better than the conventional method based on linear regression. MAPE of the proposed method is 9.0973%, while MAPE of the conventional method is 9.4533%.

Keywords:Very short-term load forecasting, office building, pattern, temperature

Full Paper.pdf
Copyright © 2016 International Journal of Smart Grid and Clean Energy, All Rights Reserved