SIU Logo DSpace
 

SIU Knowledge Bank >
The SIU Libraries >
Reprints of Faculty & Student Publications >
School of Technology Publications (SoT) >

Please use this identifier to cite or link to this item: http://dspace.siu.ac.th/handle/1532/654

Title: An accuracy assessment of an empirical sine model, a novel sine model and an artificial neural network model for forecasting illuminance/irradiance on horizontal plane of all sky types at Mahasarakham, Thailand
Authors: Pattanasethanon, Singthong
Soponronnarit, Somchart
Keywords: Correlation (Statistics)
Artificial neural network
Mean bias deviation
Root-mean-square-deviation
Coefficient of correlation
Alternative model
Issue Date: 2008
Type: Preprint
Series/Report no.: REP503 2008;c.1
REP503 2008;c.2
Abstract: The results of a study on all sky modeling and forecasting daylight availability for the tropical climate found in the central region of the northeastern part of Thailand (16°14′ N, 103°15′ E) is presented. The required components of sky quantities, namely, global and diffuse horizontal irradiance and global horizontal illuminance for saving energy used in buildings are estimated. The empirical sinusoidal models are validated. A and B values of the empirical sinusoidal model for all sky conditions are determined and developed to become a form of the sky conditions. In addition, a novel sinusoidal model, which consists of polynomial or exponential functions, is validated. A and B values of the empirical sinusoidal model for all sky conditions are determined and developed to become a new function in the polynomial or exponential form of the sky conditions. Novelettes, an artificial intelligent agent, namely, artificial neural network (ANN) model is also identified. Back propagation learning algorithms were used in the networks. Moreover, a one year data set and a next half year data set were used in order to train and test the neural network, respectively. Observation results from one years round data indicate that luminosity and energy from the sky on horizontal in the area around Mahasarakham are frequently brighter than those of Bangkok. The accuracy of the validated model is determined in terms of the mean bias deviation (MBD), the root-mean-square-deviation (RMSD) and the coefficient of correlation (R2) values. A comparison of the estimated solar irradiation values and the observed values revealed a small error slide in the empirical sinusoidal model as well. In addition, some results of the sky quantity forecast by the ANN model indicate that the ANN model is more accurate than the empirical models and the novel sinusoidal models. This study confirms the ability of the ANN to predict highly accurate solar radiance/illuminance values. We believe that the ANN model is suitable as an alternative model for forecasting the sky quantities.
Description: Energy Conversion and Management. Vol.49 No.8 (2008) p.1999-2005.
URI: http://dspace.siu.ac.th/handle/1532/654
Appears in Collections:School of Technology Publications (SoT)

Files in This Item:

File Description SizeFormat
REP503 2008.pdf1.99 MBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

 

Valid XHTML 1.0!
Powered by DSpace
Copyright © 2009-2012 Shinawatra University : Fostering Innovation
Visit our Website for more information
Contact SIU Libraries