Forecasting Electricity Demand Using a Hybrid Statistical Model

DOI

10.2991/assehr.k.201214.267How to use a DOI?

Keywords

ARIMA, ANN, electricity

Abstract

In this study, a total of four models were utilized, the pure ARIMA and ANN models and two different methodologies that were used to combine ARIMA (Autoregressive Integrated Moving Average) and ANN (Artificial Neural Network) models, an additive methodology and a multiplicative methodology, to model the electricity consumption data of Cagayan de Oro City. Results from the evaluation of all four models, the two pure and two hybrid models, have shown that with a MAPE of 0.7411 and MSE of 9.43 × 1011, ANN was the best model that fit the electricity consumption data of Cagayan de Oro City. Out of the two hybrid models, the additive ARIMA-ANN hybrid model performed better than the multiplicative model, having a MAPE of 1.3896 and a MSE value of 1.67 × 1012 compared to evaluation values of MAPE = 1.4258 and MSE = 1.68 × 1012. Out of the four, ARIMA performed the least, with a MAPE of 1.4334 and MSE of 1.70 × 1012. Results from forecasting using ANN have shown that in three years, the electricity consumption data of the city will increase, with an average monthly growth rate of 0.3123%.

Copyright

© 2020, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Arvin Paul Sumobay
PY  - 2020
DA  - 2020/12/15
TI  - Forecasting Electricity Demand Using a Hybrid Statistical Model
BT  - Proceedings of the 1 st International Conference on Information Technology and Education (ICITE 2020)
PB  - Atlantis Press
SP  - 395
EP  - 402
SN  - 2352-5398
UR  - https://doi.org/10.2991/assehr.k.201214.267
DO  - 10.2991/assehr.k.201214.267
ID  - Sumobay2020
ER  -