Volume 4, Issue 3 (12-2014)                   JEM 2014, 4(3): 38-47 | Back to browse issues page

XML Persian Abstract Print

Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:

Ameri M, Hadipoor M. Evaluation of Artificial Neural Network Performance to Predict Daily Solar Radiation in Iran. JEM. 2014; 4 (3) :38-47
URL: http://energy.kashanu.ac.ir/article-1-219-en.html
Shahid Bahonar University
Iran has an average of 5.5 KWh per square meter solar radiation and 300 sunny days per year on 90% of the land. Regarding this amount of solar radiation and the necessity for solar potential zoning for better efficiencies, drawing solar potential maps is essential. In this study, the monthly data of 39 synoptic of Iran meteorological stations over years (1991-2000) has been used as the input data to the MATLAB software and artificial neural network (ANN). In the ANN, a multi-layer feed forward model is used. After applying the input data to the network with desired architecture, in output layer the solar radiation is predicted. The solar radiation anticipated by ANN is highly in accordance with meteorological data so that the final correlation coefficient is 0.96, depicting the great accuracy of the data derived from the software. By selecting the predicted data of ANN as input to ArcGIS software, the annual solar potential map of Iran is obtained.
Type of Study: Research | Subject: Mechanical Engineering

Add your comments about this article : Your username or Email:

© 2020 All Rights Reserved | Energy Engineering & Management

Designed & Developed by : Yektaweb