Short-term Electricity Load Forecasting in Thailand: an Analysis on Different Input Variables

2018 2nd International Conference on Power and Energy Engineering

Su, Wutyi Hnin^1 ; Chawalit, Jeenanunta^1

School of Management Technology, Sirindhorn International Institute of Technology, Thammasat University, Pathum Thani

12120, Thailand^1

关键词: Accurate performance;    Data preprocessing;    Electricity Generating Authority of Thailand;    Filtering technique;    Mean absolute percentage error;    Short term load forecasting;    Short-term electricity load forecasting;    Support vector regression models;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/192/1/012040/pdf
DOI  :  10.1088/1755-1315/192/1/012040

来源: IOP

【 摘 要 】

This paper suggests a support vector regression model to make short-term load forecasting in Thailand by different training inputs. The primary objective of this paper is to describe the importance of data pre-processing and the external factors for accurate forecasting. The Electricity Generating Authority of Thailand (EGAT) provides the half-hourly electricity load demand. For numerical analysis, a dataset of net peak load of Thailand for a period of weeks from January 2016 to December 2017 is selected. The historical load demand is filtered for each day by Local regression filtering technique. After filtering the data, the effectiveness of input variables is important for accurate performance. Mean absolute percentage error (MAPE) is used to evaluate the model performance. By comparing the three models, model which considerate the temperature and seasonal factors enhances the model performance.