USP Electronic Research Repository

Wind Speed Forecasting using Regression, Time Series and Neural Network Models: A Case Study of Kiribati

Arzu, Arieni and Kutty, Saiyad Shamsher and Ahmed, Mohammed R. and Khan, Mohammad G.M. (2018) Wind Speed Forecasting using Regression, Time Series and Neural Network Models: A Case Study of Kiribati. [Conference Proceedings]

[thumbnail of Wind_Analysis_(Arieni,_AFMS_2018).pdf] PDF
Restricted to Repository staff only

Download (295kB) | Request a copy

Abstract

There is an increase in demand for renewable sources of energy due to apprehensions about climate change, increase in the energy demand and unpredictability of the prices and supply of fossil fuels. Wind energy is one of the world’s fastest growing sources of energy. As a result of the stochastic behavior of wind, the demand for accurate wind forecasting has become imperative to reduce the risk of uncertainty. In this paper, the wind speed data are modelled and forecasted using three forecasting techniques: Multiple Linear Regression (MLR), Autoregressive Integrated Moving Average (ARIMA) and Artificial Neural Network (ANN). To test these models for
wind speed forecasting, daily wind speed, pressure, relative humidity and temperature data for the period of September 2012 to September 2013 for Abaiang in Kiribati were used in this work. The performance of the models was evaluated using four measures: root mean square error, mean absolute error, mean absolute percentage error and coefficient of determination (R2). The optimum model was also compared to a benchmark technique, persistence method. The empirical results reveal that the proposed model using Artificial Neural Network is more efficient and accurate in forecasting wind speed in comparison
to the regression and time series models.

Item Type: Conference Proceedings
Subjects: Q Science > Q Science (General)
Divisions: Faculty of Science, Technology and Environment (FSTE) > School of Computing, Information and Mathematical Sciences
Depositing User: Komal Devi
Date Deposited: 05 Feb 2024 23:19
Last Modified: 05 Feb 2024 23:19
URI: https://repository.usp.ac.fj/id/eprint/11331

Actions (login required)

View Item View Item