Open AccessA Hybrid Modeling of Long-Term Electricity Consumption Forecasting Based on ARIMA and ANN: The Case of Thailand with Projection
Author(s) -
Nattapon Jaisumroum,
Jirarat Teeravaraprug
Publication year - 2018
Publication title -
international journal of engineering and technology
Language(s) - English
Resource type - Journals
ISSN - 2227-524X
DOI - 10.14419/ijet.v7i2.28.12875
Subject(s) - autoregressive integrated moving average , mean absolute percentage error , mean squared error , econometrics , statistics , artificial neural network , gross domestic product , per capita , approximation error , term (time) , electricity , population , mathematics , computer science , time series , engineering , economics , artificial intelligence , physics , quantum mechanics , sociology , electrical engineering , demography , economic growth
Modeling and forecasting of electricity consumption can provide reliable guidance for power operation and planning in developing countries such as Thailand. In this study, formulates the effects of two different historical data type is modeled by auto regressive integrated moving averaged (ARIMA) and artificial neural network (ANN) based on population and gross domestic product per capita (GDP). The derived model is validated by various statistical approaches such as the determination coefficient. Additionally, the performances of the derived model are assessed using mean absolute percentage error (MAPE), root mean square error (RMSE) and mean absolute error (MAE). Three scenarios are used for forecasting Thailand’s electricity consumption in 2011 – 2015. The simulation results are validated by actual data sets observed from 1993 to 2010. Empirical results showed that the proposed method has higher accuracy compared to single ARIMA and artificial intelligence based models.