Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
An artificial neural network model for forecasting daily global solar radiation in Ibadan, Nigeria
Solar radiation, the primary driver for many physical, chemical and biological processes on the earth's surface is considered the most indispensable parameter in the performance prediction of solar power systems. In this study, an artificial neural network (ANN) model was developed for predicting mi...
Saved in:
| Format: | Article |
|---|---|
| Published: |
2009
|
| Tags: |
No Tags, Be the first to tag this record!
|
MARC
| LEADER | 00000njm a2000000a 4500 | ||
|---|---|---|---|
| 001 | oai:repository.ui.edu.ng:123456789/1916 | ||
| 042 | |a dc | ||
| 720 | |a Fadare, D. A. |e author | ||
| 720 | |a Olugasa, T. T. |e author | ||
| 260 | |c 2009 | ||
| 520 | |a Solar radiation, the primary driver for many physical, chemical and biological processes on the earth's surface is considered the most indispensable parameter in the performance prediction of solar power systems. In this study, an artificial neural network (ANN) model was developed for predicting missing solar radiation data for Ibadan (Lat. 7.43°N; Long. 3.9°E; Alt. 227.2m), Nigeria. This study utilized daily solar radiation data for the period of 1984 to 2007 (24 years) from a meteorological station in Ibadan. The ANN model was designed using the Matlab® Neural Network Toolbox and five different structures of the model were investigated. Structure 1 utilized solar radiation data for 5 days to predict the next 25 days expected data; structure 2 utilized data for 10 days to predict the next 20 days; structure 3 used data for 15 days to predict succeeding 15 days; structure 4 used data for 25 days to predict next 5 days data; structure 5 used data for 5 days to predict the next 1 day solar radiation. The different structures were trained by using solar radiation data for 22 years and one year and the prediction accuracies were evaluated using the solar radiation values for year 2007. Results showed that structure 5 with correlation coefficient of 0.73 and 0.79 when trained with 22 years and 1 year, respectively gave the best prediction performance. Thus, indicating the suitability of structure 5 for prediction of solar radiation missing data. | ||
| 024 | 8 | |a ui_art_fadare_artificial_2009 | |
| 024 | 8 | |a Global Journal of Engineering and Technology 2(2), pp. 211-221 | |
| 024 | 8 | |a http://ir.library.ui.edu.ng/handle/123456789/1916 | |
| 245 | 0 | 0 | |a An artificial neural network model for forecasting daily global solar radiation in Ibadan, Nigeria |