Physical and hybrid methods comparison for the day ahead PV output power forecast

An accurate forecast of the exploitable energy from Renewable Energy Sources, provided 24 h in advance, is becoming more and more important in the context of the smart grids, both for their stability issues and the reliability of the bidding markets. This work presents a comparison of the PV output power day-ahead forecasts performed by deterministic and stochastic models aiming to find out the best performance conditions. In particular, we have compared the results of two deterministic models, based on three and five parameters electric equivalent circuit, and a hybrid method based on artificial neural network. The forecasts are evaluated against real data measured for one year in an existing PV plant located at SolarTechlab in Milan, Italy. In general, there is no significant difference between the two deterministic models, being the three-parameter approach slightly more accurate (NMAE three-parameter 8.5% vs. NMAE five-parameter 9.0%). The artificial neural network, combined with clear sky solar radiation, generally achieves the best forecasting results (NMAE 5.6%) and only few days of training are necessary to provide accurate forecasts.

[1]  Christian Borgelt,et al.  Introduction to Neural Networks , 2016 .

[2]  William A. Beckman,et al.  Improvement and validation of a model for photovoltaic array performance , 2006 .

[3]  Rajesh Gupta,et al.  Parameter extraction for dynamic PV thermal model using particle swarm optimisation , 2016 .

[4]  Marco Mussetta,et al.  Day-ahead PV Power Forecast by Hybrid ANN Compared to the Five Parameters Model Estimated by Particle Filter Algorithm , 2016, ICANN.

[5]  H. Hottel.  A simple model for estimating the transmittance of direct solar radiation through clear atmospheres , 1976 .

[6]  Francesco Grimaccia,et al.  Hybrid Predictive Models for Accurate Forecasting in PV Systems , 2013 .

[7]  Ignacio J. Ramirez-Rosado,et al.  Short-Term Forecasting Models for Photovoltaic Plants: Analytical versus Soft-Computing Techniques , 2013 .

[8]  Francesco Grimaccia,et al.  A Physical Hybrid Artificial Neural Network for Short Term Forecasting of PV Plant Power Output , 2015 .

[9]  Francesco Grimaccia,et al.  Hybrid model for hourly forecast of photovoltaic and wind power , 2013, 2013 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE).

[10]  R. Urraca,et al.  Review of photovoltaic power forecasting , 2016 .

[11]  Francesco Grimaccia,et al.  Hybrid model analysis and validation for PV energy production forecasting , 2014, 2014 International Joint Conference on Neural Networks (IJCNN).

[12]  Leah L. Rogers,et al.  Solving Problems in Environmental Engineering and Geosciences with Artificial Neural Networks , 1996 .

[13]  Jan Kleissl,et al.  Solar Energy Forecasting and Resource Assessment , 2013 .

[14]  Ali Naci Celik.  Artificial neural network modelling and experimental verification of the operating current of mono-crystalline photovoltaic modules , 2011 .

[15]  Víctor Manuel Fernandes Mendes,et al.  Simulation of a Solar Cell considering Single-Diode Equivalent Circuit Model , 2011 .

[16]  F.A. Farret,et al.  New methodology to determinate photovoltaic parameters of solar panels , 2011, XI Brazilian Power Electronics Conference.

[17]  Francesco Grimaccia,et al.  Neural forecasting of the day-ahead hourly power curve of a photovoltaic plant , 2016, 2016 International Joint Conference on Neural Networks (IJCNN).

[18]  Giles M. Foody,et al.  An evaluation of some factors affecting the accuracy of classification by an artificial neural network , 1997 .

[19]  J. Nelson.  The physics of solar cells , 2003 .

[20]  A. Mellit,et al.  A 24-h forecast of solar irradiance using artificial neural network: Application for performance prediction of a grid-connected PV plant at Trieste, Italy , 2010 .

[21]  A. Dolara,et al.  Comparison of different physical models for PV power output prediction , 2015 .

[22]  David Haussler,et al.  What Size Net Gives Valid Generalization? , 1989, Neural Computation.

[23]  Francesco Grimaccia,et al.  Analysis and validation of 24 hours ahead neural network forecasting of photovoltaic output power , 2017, Math. Comput. Simul..

[24]  Antonio Luque,et al.  Handbook of photovoltaic science and engineering , 2011 .

[25]  Oliver Kramer,et al.  Comparing support vector regression for PV power forecasting to a physical modeling approach using measurement, numerical weather prediction, and cloud motion data , 2016 .

[26]  F. Hocaoglu,et al.  Hourly solar radiation forecasting using optimal coefficient 2-D linear filters and feed-forward neural networks , 2008 .

[27]  S. Leva,et al.  PV hourly day-ahead power forecasting in a micro grid context , 2016, 2016 IEEE 16th International Conference on Environment and Electrical Engineering (EEEIC).

[28]  E. Izgi,et al.  Short–mid-term solar power prediction by using artificial neural networks , 2012 .

[29]  Anthony J. Jakeman,et al.  Artificial Intelligence techniques: An introduction to their use for modelling environmental systems , 2008, Math. Comput. Simul..

[30]  Michael Y. Hu,et al.  Forecasting with artificial neural networks: The state of the art , 1997 .

[31]  S. Arulselvi,et al.  Photovoltaic modeling and its characteristics , 2011, 2011 International Conference on Emerging Trends in Electrical and Computer Technology.