IWANN'99

Improved Generation of Hourly Solar Irradiation Artificial Series Using Neural Networks

L. Hontoria+
hontoria@ujaen.es
J. Riesco*
jacobo@geocities.com
P. Zufiria*
pzz@mat.upm.es
J. Aguilera+
aguilera@ujaen.es
+ Grupo Jaén de Técnica Aplicada.
Departamento de Electrónica.
Universidad de Jaén.
Avda. Madrid, 35, 23071 Jaén. Spain.
* Grupo de Redes Neuronales.
Departamento de Matemática Aplicada a las Tecnologías de la Información.
E.T.S. Ingenieros de Telecomunicación.
Universidad Politécnica de Madrid.

Ciudad Universitaria s/n, 28040 Madrid. Spain.

International Conference on Engineering Applications of Neural Networks (EANN'99)
Warsaw (Poland). September 13-15, 1999.

ABSTRACT
This paper presents a new neural network approach for the generation of synthetic hourly irradiation series, a relevant problem in the photo-voltaic field. The neural model employed is the well known Multi-Layer Perceptron (MLP) paradigm, in a feedback architecture, using a record of historical values for the supervised network training. The method is based on the MLP ability to extract, from a sufficiently general training set, the existing relationships between variables whose interdependence is unknown a priori. Simulation results are compared to other methods, and show that the generated values follow the average tendency of the real values. Though the method has been developed using data values from Madrid, it can be generalised to any location. Even more, the proposed methodology is of general applicability to the estimation of highly complex temporal series.


Related files:
[Paper (PDF)] eann99.pdf Paper (Adobe Acrobat Portable Document Format PDF)
[Poster (PDF)] eann99pp.pdf Presentation Slides (PDF)


September-1999
1