Solar irradiation prediction with machine learning: Forecasting models selection method depending on weather variability

Abstract Eleven statistical and machine learning tools are analyzed and applied to hourly solar irradiation forecasting for time horizon from 1 to 6 h. A methodology is presented to select the best and most reliable forecasting model according to the meteorological variability of the site. To make the conclusions more universal, solar data collected in three sites with low, medium and high meteorological variabilities are used: Ajaccio, Tilos and Odeillo. The datasets variability is evaluated using the mean absolute log return value. The models were compared in term of normalized root mean square error, mean absolute error and skill score. The most efficient models are selected for each variability and temporal horizon: for the weak variability, auto-regressive moving average and multi-layer perceptron are the most efficient, for a medium variability, auto-regressive moving average and bagged regression tree are the best predictors and for a high one, only more complex methods can be used efficiently, bagged regression tree and the random forest approach.

[1]  Adel Mellit,et al.  Artificial Intelligence technique for modelling and forecasting of solar radiation data: a review , 2008, Int. J. Artif. Intell. Soft Comput..

[2]  R. Tibshirani,et al.  Generalized additive models for medical research , 1986, Statistical methods in medical research.

[3]  P. Ineichen.  Comparison of eight clear sky broadband models against 16 independent data banks , 2006 .

[4]  He Jiang,et al.  Global horizontal radiation forecast using forward regression on a quadratic kernel support vector machine: Case study of the Tibet Autonomous Region in China , 2017 .

[5]  Yi Zheng,et al.  Mutual information for evaluating renewable power penetration impacts in a distributed generation system , 2017 .

[6]  William R. Burrows,et al.  CART Regression Models for Predicting UV Radiation at the Ground in the Presence of Cloud and Other Environmental Factors , 1997 .

[7]  Rob J Hyndman,et al.  25 years of time series forecasting , 2006 .

[8]  Soteris A. Kalogirou,et al.  Machine learning methods for solar radiation forecasting: A review , 2017 .

[9]  Gilles Notton,et al.  Forecasting method for global radiation time series without training phase: Comparison with other well-known prediction methodologies , 2017 .

[10]  Cyril Voyant,et al.  Forecasting of preprocessed daily solar radiation time series using neural networks , 2010 .

[11]  John Boland,et al.  Short term solar radiation forecasting: Island versus continental sites , 2016 .

[12]  T. Hoff,et al.  Short-term irradiance variability: Preliminary estimation of station pair correlation as a function of distance , 2012 .

[13]  R. Kuhlemann,et al.  Rethinking satellite-based solar irradiance modelling: The SOLIS clear-sky module , 2004 .

[14]  Nasser Mozayani,et al.  Mutual Information Based Input Variable Selection Algorithm and Wavelet Neural Network for Time Series Prediction , 2008, ICANN.

[15]  P. Ineichen.  A broadband simplified version of the Solis clear sky model , 2008 .

[16]  M. Diagne,et al.  Review of solar irradiance forecasting methods and a proposition for small-scale insular grids , 2013 .

[17]  Viorel Badescu,et al.  Modeling Solar Radiation at the Earth's Surface: Recent Advances , 2014 .

[18]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[19]  Li Wang,et al.  Uncertainty analysis of an integrated energy system based on information theory , 2017 .

[20]  Sancho Salcedo-Sanz,et al.  Local models-based regression trees for very short-term wind speed prediction , 2015 .

[21]  Tamer Khatib,et al.  A novel hybrid model for hourly global solar radiation prediction using random forests technique and firefly algorithm , 2017 .

[22]  Carl E. Rasmussen,et al.  Gaussian processes for machine learning , 2005, Adaptive computation and machine learning.

[23]  Kurt Hornik,et al.  Multilayer feedforward networks are universal approximators , 1989, Neural Networks.

[24]  M. Iqbal.  An introduction to solar radiation , 1983 .

[25]  Cyril Voyant,et al.  Statistical parameters as a means to a priori assess the accuracy of solar forecasting models , 2015 .

[26]  Leo Breiman,et al.  Bagging Predictors , 1996, Machine Learning.

[27]  G. De’ath.  Boosted trees for ecological modeling and prediction. , 2007, Ecology.

[28]  Lalit Mohan Saini,et al.  Solar energy prediction using linear and non-linear regularization models: A study on AMS (American Meteorological Society) 2013–14 Solar Energy Prediction Contest , 2014 .

[29]  Soteris A. Kalogirou,et al.  Applications of artificial neural-networks for energy systems , 2000 .

[30]  Carlos F.M. Coimbra,et al.  Assessment of machine learning techniques for deterministic and probabilistic intra-hour solar forecasts , 2018, Renewable Energy.

[31]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[32]  Jing Huang,et al.  A semi-empirical approach using gradient boosting and k-nearest neighbors regression for GEFCom2014 probabilistic solar power forecasting , 2016 .

[33]  Fernando Luiz Cyrino Oliveira,et al.  Forecasting mid-long term electric energy consumption through bagging ARIMA and exponential smoothing methods , 2018 .

[34]  T. Soubdhan,et al.  A benchmarking of machine learning techniques for solar radiation forecasting in an insular context , 2015 .

[35]  Henrik Madsen,et al.  Multi-site solar power forecasting using gradient boosted regression trees , 2017 .