1. A REVIEW OF SOLAR IRRADIANCE PREDICTION
TECHNIQUES
Martín, L.; Zarzalejo, L. F.; Polo, J.; Espinar, B. & Ramírez, L
SOLAR RESOURCE KNOWLEDGE MANAGEMENT
TASK No. 36
IEA Solar Heating & Cooling Programme
CIEMAT WORKING GROUP (SPAIN)
SUBTASK A: Standard Qualification For Solar Resource Products
6-7 July 2006 Denver, Colorado
2. SOLAR PREDICTION OVERVIEW
• Solar Energy:
– Dynamic in the atmosphere the oceans and in general life on earth.
– Solar water heating, water detoxification, water desalinization, electric power energy
generation from solar thermal power and photovoltaic energy, agricultural applications ….
• Need to characterize and predict incoming solar radiation to be used
as a energetic resource.
• Prediction General Techniques
1. Numerical Weather Predictions Models
2. Statistical Prediction
• Forecasting Horizon
– Nowcasting
– Short Term
– Medium Term
– Long Term
3. MOS SOLAR PREDICTION – SHORT TERM
Differents Works from 80s:
John S. Jensenius& Gerald F. Cotton, 1981:
The developmentand testing of automated Solar energy forecasts based on the model output statistics (MOS)
technique
1st Workshop On Terrestrial SolarResource Forecasting and
on the Use on Satellites for Terrestrial Solar Resource Assesssment, Newark, 1981, Am. Sol. En. Soc.
New appraches using sky cover product from wheather prediction
centers:
GHI
= g ( SK )
GHI clear − sky
4. SATELLITE SOLAR PREDICTION
Annette Hammer, Detlev Heinemann, Carster Hoyer, Elke Lorenz. Satellite based short-term
forecasting of solar irradiance - comparison of methods and error analysis. 2000.
5. SIGNAL ANALISYS AND ARTIFICIAL INTELIGENT
APPROACHES
Cao S, Cao J. Forecast of solar irradiance using recurrent neural networks combined with wavelet
analysis. Applied Thermal Engineering 2005 Feb;25(2-3):161-72.
Signal Analysis Time-Frecuency (Scale) with Wavelet Transform
Señal Original Normalizada / Año 2001
Discrete Wavelet Transform:
Approximation A3 Detail D1
1 0.8 0.4
0.6 0.2
0.8
Signals Filtered: 0.4 0
High Frecuency (Detail)
0.6 0.2 -0.2
0 -0.4
Low Frecuency (Aproximation)
0.4
-0.2 -0.6
0 50 100 150 200 250 300 350 400 0 50 100 150 200 250 300 350 40
0.2
Detail D2 Detail D3
0.4 0.3
0.2
0
0.2
0.1
0
-0.2 0
-0.1
-0.2
-0.2
-0.4 -0.3
0 50 100 150 200 250 300 350 400
-0.4 -0.4
0 50 100 150 200 250 300 350 400 0 50 100 150 200 250 300 350 40
Prediction with Artificial Neural Networks
6. FUTURE WORKS
• Wavelet analysis and NN with Normalized data (Kt).
• Use other NN architectures like Self-Organized Features
Maps (SOFMs).
• Use network surface irradiance data forecasted from NWP
from European Centre Medium Weather Forecasting
(ECMWF) as a new parameter in NN.
• Wavelet and temporal series technique.
• Motion estimation with segmentation techniques in
satellite images.
• Med-Long Term Prediction: EOF Analysis analysis to
relate different atmospheric oscillation patterns, NAO
(North Atlantic Oscillation), ENSO (El Niño-Southern
Oscillation),… with expected solar irradiance.
7. FUTURE WORKS
• Wavelet analysis and NN with Normalized data (Kt).
• Use other NN architectures like Self-Organized Features
Maps (SOFMs).
• Use network surface irradiance data forecasted from NWP
from European Centre Medium Weather Forecasting
(ECMWF) as a new parameter in NN.
• Wavelet and temporal series technique.
• Motion estimation with segmentation techniques in
satellite images.
• Med-Long Term Prediction: EOF Analysis analysis to
relate different atmospheric oscillation patterns, NAO
(North Atlantic Oscillation), ENSO (El Niño-Southern
Oscillation),… with expected solar irradiance.