The Multi-Layered Perceptrons Neural Networks for the Prediction of Daily Solar Radiation

The Multi-Layered Perceptron (MLP) Neural networks have been very successful in a number of signal processing applications. In this work we have studied the possibilities and the met difficulties in the application of the MLP neural networks for the prediction of daily solar radiation data. We have used the Polack-Ribière algorithm for training the neural networks. A comparison, in term of the statistical indicators, with a linear model most used in literature, is also performed, and the obtained results show that the neural networks are more efficient and gave the best results.

Authors:

• Radouane Iqdour
• Abdelouhab Zeroual

Keywords:

• Daily solar radiation
• Prediction
• MLP neural networks
• linear model

References:

[1] H. Bouhadou, MM. Hassani, A. Zeroual, Wilkinson AJ, "Stochastic
Simulation of weather data using higher order statistics", Renewable
Energy , 12(1), 1997, pp. 21-37.
[2] S. Safi, A. Zeroual, "MA system identification using higher order
cumulants: application to modeling solar radiation", Journal of
Statistical Computation and Simulation, Vol.72 (7) , 2002, pp. 533-548,.
[3] S. Safi, A. Zeroual, MM. Hassani, "Prediction of global daily solar
radiation using higher order statistics", Renewable Energy, 27, 2002, pp.
647-666.
[4] S. Safi, A. Zeroual, "Modelling solar data using high order statistics"
A.M.S.E. Advances in Modelling & Analysis, Vol. 6, N┬░1, 2, Advances
D-2001, pp. 1-16..
[5] M. T. Hagan, H. B. Demuth, and M. H. Beale, Neural Network Design,
Boston, MA: PWS Publishing, 1996.
[6] J.J. Hopfield "Neural Networks and Physical Systems with Emergent
Collective Computational Abilities", Proceeding of the Natl. Acd. Sci.,
n┬░79, 1982, pp. 25554-2558.
[7] Hornik K., Stinchcombe M., White H. "Multilayer Feedforward
Networks are universal Appoximators" Neural Networkd, n┬░2, pp .359-
366.
[8] Soteris A. Kalogirou "Artificial neural networks in renewable energy
systems applications: a review" Renewable and Sustainable Energy
Reviews 5, 2001, 373-401.
[9] R. Iqdour, A. Zeroual, "Prédiction de l-irradiation journalière ├á l-aide
des réseaux de neurones MLP", International Conference on
Approximation Methods and numerical Modeling in environment and
Natural Resources MAMERN 2005, May 9-11, 2005 Oujda - Morocco.
[10] R. Iqdour, A. Zeroual, "An application of the MLP neural networks to
the prediction of daily solar radiation", IVème Conférence
Internationale en Recherche Opérationnelle Théorie et Applications
CIRO 05, 22-26 Mai 2005 Marrakech - Morocco.
[11] E. Polak, Computational Methods in Optimisation: a Unified Approach,
Editions Academic Press.
[12] J. Dayhoff Neural Networks Architectures, Editions Van Norstrand
Reynold.
[13] B. Widrow, "Adaline and madaline", Proceedings of the 1st
International Conference on Neural Networks, pp.143-158.
[14] T. Tollenaere, "Fast Adaptive Backpropagation with good Scaling
Properties" Neural Networks, n┬░3, pp.561-573.
[15] J.S. Armstrong, F. Collopy, "Error Measures for Generalzing About
Forecasting Methods: Empirical Comparison ", International Journal of
Forecast in, n┬░8, pp.69-80.
[16] D.R.Hush, B.G. Horne, "Progress in Supervised Neural Networks",
IEEE Signal Processing Magazine, n┬░10, pp.8-39.
[17] G.E.P. Box, G.M. Jenkins, Time Series Analysis, Forecasting and
Control, Editions Holden Day, 1970.