Deep Learning for Big Data Time Series Forecasting Applied to Solar Power

Accurate solar energy prediction is required for the integration of solar power into the electricity grid, to ensure reliable electricity supply, while reducing pollution. In this paper we propose a new approach based on deep learning for the task of solar photovoltaic power forecasting for the next day. We firstly evaluate the performance of the proposed algorithm using Australian solar photovoltaic data for two years. Next, we compare its performance with two other advanced methods for forecasting recently published in the literature. In particular, a forecasting algorithm based on similarity of sequences of patterns and a neural network as a reference method for solar power forecasting. Finally, the suitability of all methods to deal with big data time series is analyzed by means of a scalability study, showing the deep learning promising results for accurate solar power forecasting.

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

[2]  Zheng Wang,et al.  Solar power prediction using weather type pair patterns , 2017, 2017 International Joint Conference on Neural Networks (IJCNN).

[3]  Zheng Wang,et al.  Solar Power Forecasting Using Pattern Sequences , 2017, ICANN.

[4]  Luís Torgo,et al.  Ensembles for Time Series Forecasting , 2014, ACML.

[5]  Frederico G. Guimarães,et al.  A GPU deep learning metaheuristic based model for time series forecasting , 2017 .

[6]  Carlos F.M. Coimbra,et al.  Short-term reforecasting of power output from a 48 MWe solar PV plant , 2015 .

[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]  undefined Manoël Rekinger,et al.  Global Market Outlook for Solar Power 2015-2019 , 2014 .

[9]  Luís Torgo,et al.  Arbitrated Ensemble for Time Series Forecasting , 2017, ECML/PKDD.

[10]  Yitao Liu,et al.  Deterministic and probabilistic forecasting of photovoltaic power based on deep convolutional neural network , 2017 .

[11]  Thomas Reindl,et al.  A novel hybrid approach based on self-organizing maps, support vector regression and particle swarm optimization to forecast solar irradiance , 2015 .

[12]  Zheng Wang,et al.  Solar power prediction with data source weighted nearest neighbors , 2017, 2017 International Joint Conference on Neural Networks (IJCNN).

[13]  Francisco Martinez Alvarez,et al.  Energy Time Series Forecasting Based on Pattern Sequence Similarity , 2011, IEEE Transactions on Knowledge and Data Engineering.

[14]  Irena Koprinska,et al.  2D-interval forecasts for solar power production , 2015 .

[15]  Andreas Kamilaris,et al.  Deep learning in agriculture: A survey , 2018, Comput. Electron. Agric..

[16]  Igor V. Tetko,et al.  Data modelling with neural networks: Advantages and limitations , 1997, J. Comput. Aided Mol. Des..

[17]  Alicia Troncoso Lora,et al.  Deep Learning-Based Approach for Time Series Forecasting with Application to Electricity Load , 2017, IWINAC.