Short-term load Forecasting Using Artificial Neural Network for Ban-Nathone Substation, Salavane Province, Lao PDR

Authors

  • Lamngeun Sounaphan

Keywords:

short-term load forecast, artificial neural networks, back-propagation learning, coefficient of decision

Abstract

This research presented short-term load forecasting for Ban-Nathone substation ,Salavane province , Lao PDR by using Artificial Neural Network (ANN) consisted of 1 and 2 hidden layer model with back propagation supervised learning algorithm based on the analysis of associates among daily peak load, highest-lowest temperature, weather, and characteristics of calendar days. The forecasting result was acceptable for the ANN with 1 hidden layer for it could obtain the mean absolute percentage error (MAPE) of 2.82%, which was more accurate than that of 2 hidden layers using the same and different numbers of neural with mean absolute percentage error of 3.05%, and 3.06%, respectively. This research used the daily peak load data in 2014 from the industrial area (JV06: feeder Salavane2), and the weather information in 2014 consisted of the highest temperature, average humidity, rainfall intensity, and the considered time interval for load demand determination. These factors had been analyzed using
scatter diagram and linear regression equation (Linear Regression: R2).

Downloads

Published

2017-10-30

How to Cite

Sounaphan, L. (2017). Short-term load Forecasting Using Artificial Neural Network for Ban-Nathone Substation, Salavane Province, Lao PDR. Creative Science, 9(3), 607–616. Retrieved from https://ph01.tci-thaijo.org/index.php/snru_journal/article/view/102451