Improving solar forecasting using Deep Learning and Portfolio Theory integration

Solar energy has been consolidated as one of the main renewable energy sources capable of contributing to supply global energy demand. However, the solar resource has intermittent feature in electricity production, making it difficult to manage the electrical system. Hence, we propose the application of Deep Learning (DL), one of the emerging themes in the field of Artificial Intelligence (AI), as a solar predictor. To attest its capacity, the technique is compared with other consolidated solar forecasting strategies such as Multilayer Perceptron, Radial Base Function and Support Vector Regression. Additionally, integration of AI methods in a new adaptive topology based on the Portfolio Theory (PT) is proposed hereby to improve solar forecasts. PT takes advantage of diversified forecast assets: when one of the assets shows prediction errors, these are offset by another asset. After testing with data from Spain and Brazil, results show that the Mean Absolute Percentage Error (MAPE) for predictions using DL is 6.89% and for the proposed integration (called PrevPT) is 5.36% concerning data from Spain. For the data from Brazil, MAPE for predictions using DL is 6.08% and 4.52% for PrevPT. In both cases, DL and PrevPT results are better than the other techniques being used.

[1]  K. Ponnambalam,et al.  Risk-averse stochastic programming approach for microgrid planning under uncertainty , 2017 .

[2]  Saifur Rahman,et al.  Solar irradiance forecast using aerosols measurements: A data driven approach , 2018, Solar Energy.

[3]  Ning Xu,et al.  Forecasting district-scale energy dynamics through integrating building network and long short-term memory learning algorithm , 2019, Applied Energy.

[4]  Yongxiang Huang,et al.  Intermittency study of high frequency global solar radiation sequences under a tropical climate , 2013 .

[5]  APPLYING MODERN PORTFOLIO THEORY TO PLANT ELECTRICITY PLANNING IN ALBANIA , 2015 .

[6]  Emanuele Crisostomi,et al.  Day-Ahead Hourly Forecasting of Power Generation From Photovoltaic Plants , 2018, IEEE Transactions on Sustainable Energy.

[7]  He Jiang,et al.  Intelligent optimization models based on hard-ridge penalty and RBF for forecasting global solar radiation , 2015 .

[8]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[9]  Seth Blumsack,et al.  The capacity value of optimal wind and solar portfolios , 2018 .

[10]  K. Kaba,et al.  Estimation of daily global solar radiation using deep learning model , 2018, Energy.

[11]  Nikolaos S. Thomaidis,et al.  Exploring the mean-variance portfolio optimization approach for planning wind repowering actions in Spain , 2017 .

[12]  Kok Soon Tey,et al.  Forecasting of photovoltaic power generation and model optimization: A review , 2018 .

[13]  Andrea Vitali,et al.  Bayesian deep learning based method for probabilistic forecast of day-ahead electricity prices , 2019, Applied Energy.

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

[15]  Le Zhang,et al.  Ensemble deep learning for regression and time series forecasting , 2014, 2014 IEEE Symposium on Computational Intelligence in Ensemble Learning (CIEL).

[16]  Nikos Kourentzes,et al.  Short-term solar irradiation forecasting based on Dynamic Harmonic Regression , 2015 .

[17]  Carsten Croonenbroeck,et al.  Renewable generation forecast studies – Review and good practice guidance , 2019, Renewable and Sustainable Energy Reviews.

[18]  Yugang Niu,et al.  Hourly day-ahead solar irradiance prediction using weather forecasts by LSTM , 2018 .

[19]  Adel Mellit,et al.  Prediction of daily and mean monthly global solar radiation using support vector machine in an arid climate , 2016 .

[20]  Josileudo R. Leite,et al.  Portfolio theory applied to solar and wind resources forecast , 2017 .

[21]  Mohammed Awad,et al.  Enhanced RBF neural network model for time series prediction of solar cells panel depending on climate conditions (temperature and irradiance) , 2016, Neural Computing and Applications.

[22]  Dipti Srinivasan,et al.  Reconciling solar forecasts: Sequential reconciliation , 2019, Solar Energy.

[23]  A. Selvakumar,et al.  Assessment of SVM, empirical and ANN based solar radiation prediction models with most influencing input parameters , 2017, Renewable Energy.

[24]  L. D. Monache,et al.  An analog ensemble for short-term probabilistic solar power forecast , 2015 .

[25]  Nasser M. Nasrabadi,et al.  Pattern Recognition and Machine Learning , 2006, Technometrics.

[26]  Tanveer Ahmad,et al.  Deep learning for multi-scale smart energy forecasting , 2019 .

[27]  Jin Hur,et al.  A hybrid spatio-temporal forecasting of solar generating resources for grid integration , 2019, Energy.

[28]  Adam R. Brandt,et al.  Short-term solar power forecast with deep learning: Exploring optimal input and output configuration , 2019, Solar Energy.

[29]  Ariana Moncada,et al.  Deep Learning to Forecast Solar Irradiance Using a Six-Month UTSA SkyImager Dataset , 2018, Energies.

[30]  Durga Toshniwal,et al.  Deep learning framework to forecast electricity demand , 2019, Applied Energy.

[31]  Carlos F.M. Coimbra,et al.  Sun-tracking imaging system for intra-hour DNI forecasts , 2016 .

[32]  W. V. Sark,et al.  Short-term peer-to-peer solar forecasting in a network of photovoltaic systems , 2017 .

[33]  Jui-Sheng Chou,et al.  The use of artificial intelligence combiners for modeling steel pitting risk and corrosion rate , 2017, Eng. Appl. Artif. Intell..

[34]  Bryan Maybee,et al.  Climate Policy Uncertainty and Power Generation Investments: A Real Options-CVaR Portfolio Optimization Approach☆ , 2015 .

[35]  Dazhi Yang,et al.  A universal benchmarking method for probabilistic solar irradiance forecasting , 2019, Solar Energy.

[36]  Wei Qiao,et al.  Short-term solar power prediction using a support vector machine , 2013 .

[37]  Stéphanie Monjoly,et al.  Hourly forecasting of global solar radiation based on multiscale decomposition methods: A hybrid approach , 2017 .

[38]  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 .

[39]  Paulo Cesar Marques de Carvalho,et al.  Innovative hybrid models for forecasting time series applied in wind generation based on the combination of time series models with artificial neural networks , 2018 .

[40]  Paula F. V. Ferreira,et al.  Designing electricity generation portfolios using the mean-variance approach , 2015 .