Univariate and multivariate methods for very short-term solar photovoltaic power forecasting

Abstract We consider the task of forecasting the electricity power generated by a solar PhotoVoltaic (PV) system for forecasting horizons from 5 to 60 min ahead, from previous PV power and meteorological data. We present a new method based on advanced machine learning algorithms for variable selection and prediction. The correlation based variable selection identifies a small set of informative variables that are used as inputs for an ensemble of neural networks and support vector regression algorithms to generate the predictions. We develop two types of models: univariate, that use only previous PV power data, and multivariate, that also use previous weather data, and evaluate their performance on Australian PV data for two years. The results show that the univariate models performed similarly to the multivariate models, achieving mean relative error of 4.15–9.34%. Hence, the PV power output for very short-term forecasting horizons of 5–60 min can be predicted accurately by using only previous PV power data, without weather information. The most accurate model was univariate ensemble of neural networks, predicting the PV power output separately for each step of the forecasting horizon.

[1]  Mark A. Hall,et al.  Correlation-based Feature Selection for Discrete and Numeric Class Machine Learning , 1999, ICML.

[2]  Germano C. Vasconcelos,et al.  MLP ensembles improve long term prediction accuracy over single networks , 2011 .

[3]  Francisco Martínez-Álvarez,et al.  A Survey on Data Mining Techniques Applied to Electricity-Related Time Series Forecasting , 2015 .

[4]  H. Pedro,et al.  Assessment of forecasting techniques for solar power production with no exogenous inputs , 2012 .

[5]  Chunxiang Yang,et al.  An Improved Photovoltaic Power Forecasting Model With the Assistance of Aerosol Index Data , 2015, IEEE Transactions on Sustainable Energy.

[6]  L. Wald,et al.  The method Heliosat-2 for deriving shortwave solar radiation from satellite images , 2004 .

[7]  Irena Koprinska,et al.  Combining pattern sequence similarity with neural networks for forecasting electricity demand time series , 2013, The 2013 International Joint Conference on Neural Networks (IJCNN).

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

[9]  M. Alia-Martinez,et al.  Smart baseline models for solar irradiation forecasting , 2016 .

[10]  Trevor Hastie,et al.  The Elements of Statistical Learning , 2001 .

[11]  Carlos E. Pedreira,et al.  Neural networks for short-term load forecasting: a review and evaluation , 2001 .

[12]  P. Cochat,et al.  Et al , 2008, Archives de pediatrie : organe officiel de la Societe francaise de pediatrie.

[13]  Detlev Heinemann,et al.  Regional PV power prediction for improved grid integration , 2011 .

[14]  F. Martínez-Álvarez,et al.  A survey on data mining techniques applied to energy time series forecasting , 2016 .

[15]  Francisco J. Batlles,et al.  Solar radiation forecasting in the short- and medium-term under all sky conditions , 2015 .

[16]  Bernhard Schölkopf,et al.  A tutorial on support vector regression , 2004, Stat. Comput..

[17]  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 .

[18]  J. A. Ruiz-Arias,et al.  Benchmarking of different approaches to forecast solar irradiance , 2009 .

[19]  Tiago Alessandro Espínola Ferreira,et al.  A New Intelligent System Methodology for Time Series Forecasting with Artificial Neural Networks , 2008, Neural Processing Letters.

[20]  Isabelle Guyon,et al.  An Introduction to Variable and Feature Selection , 2003, J. Mach. Learn. Res..

[21]  Mohammad Bagher Menhaj,et al.  Training feedforward networks with the Marquardt algorithm , 1994, IEEE Trans. Neural Networks.

[22]  Gunnar Rätsch,et al.  Predicting Time Series with Support Vector Machines , 1997, ICANN.

[23]  Irena Koprinska,et al.  Forecasting hourly electricity load profile using neural networks , 2014, 2014 International Joint Conference on Neural Networks (IJCNN).

[24]  S. Pelland,et al.  Solar and photovoltaic forecasting through post‐processing of the Global Environmental Multiscale numerical weather prediction model , 2013 .

[25]  L. Wald,et al.  On the clear sky model of the ESRA — European Solar Radiation Atlas — with respect to the heliosat method , 2000 .

[26]  Irena Koprinska,et al.  Correlation and instance based feature selection for electricity load forecasting , 2015, Knowl. Based Syst..

[27]  Irena Koprinska,et al.  Very short-term electricity load demand forecasting using support vector regression , 2009, 2009 International Joint Conference on Neural Networks.

[28]  Chao-Ming Huang,et al.  A Weather-Based Hybrid Method for 1-Day Ahead Hourly Forecasting of PV Power Output , 2014, IEEE Transactions on Sustainable Energy.

[29]  Bangyin Liu,et al.  Online 24-h solar power forecasting based on weather type classification using artificial neural network , 2011 .

[30]  Peng Wang,et al.  Forecasting Power Output of Photovoltaic Systems Based on Weather Classification and Support Vector Machines , 2011, IEEE Transactions on Industry Applications.

[31]  Dazhi Yang,et al.  Very short term irradiance forecasting using the lasso , 2015 .

[32]  Francisco J. Batlles,et al.  Solar irradiance forecasting at one-minute intervals for different sky conditions using sky camera images , 2015 .