Forecasting Power Demand Using Neural Networks Model
Shima Simsar1, Mahmood Alborzi2, Jamshid Nazemi3, Mahmood Abbasi Layegh4
1Shima Simsar, Department of IT Management, Science and Research University, Faculty of Management, Tehran, Iran.
2Mahmood Alborzi, Science and Research University, Faculty of Management, Tehran, Iran.
3Jamshid Nazemi, Science and Research University, Faculty of Management, Tehran, Iran.
4Mahmood Abbasi Layegh, Urmia University, Department of Electrical Engineering, West Azarbaijan Province, Iran.
Manuscript received on May 26, 2013. | Revised Manuscript received on June 16, 2013. | Manuscript published on June 30, 2013. | PP: 441-446 | Volume-2, Issue-5, June 2013. | Retrieval Number: E1811062513/2013©BEIESP
Open Access | Ethics and Policies | Cite
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: In recent years, by entering the competition arena, not only providing the needed electricity demand, but also reducing the cost of purchased electricity has been one of the biggest challenges of power distribution companies. Solving this challenge has lots of profits and high efficiency for these companies, this research deals with forecasting power demand using neural networks model. To test the power demand , two consecutive years in the West Azarbaijan Province have been selected as a case study. Daily consumption of electricity demand follows time series models. In this study, the daily demand for two years, temperature and humidity of each day and type of days (weekdays or weekends) have been considered. In order to fit the neural network model, the architecture of multi-layer perceptron (MLP) with back propagation learning algorithm has been used. The results indicate that data related to humidity, temperature and also weekends or off-days have an effect on prediction of electricity demand.
Keywords: Forecast, Neural network, Time series, Power demand.