A recursive ensemble model for forecasting the power output of photovoltaic systems

Abstract Solar power provides a clean and renewable energy source. However, unlike many conventional sources, Photovoltaic (PV) power generation is of high volatility and uncertainty in short terms, which creates great challenges to forecasting and balancing electricity generation with demand. This study investigates the effects of PV solar power variability and proposes a data-driven ensemble modeling technique to improve the prediction accuracy of PV power generation. Three different types of models are integrated within a recursive arithmetic average model on their stand-alone predictions. The proposed methodology is later demonstrated to be of higher accuracy by comparing its prediction performance with each stand-alone forecasting model. Several different training and testing samples have been analyzed with the proposed model. The results show that the ensemble model performs better than the other stand-alone forecasting techniques in general.

[1]  Yongtao Hao,et al.  A feature weighted support vector machine and K-nearest neighbor algorithm for stock market indices prediction , 2017, Expert Syst. Appl..

[2]  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.

[3]  Boudewijn Elsinga,et al.  An artificial neural network to assess the impact of neighbouring photovoltaic systems in power forecasting in Utrecht, the Netherlands , 2016 .

[4]  Salvador Pintos,et al.  An Optimization Methodology of Alkaline-Surfactant-Polymer Flooding Processes Using Field Scale Numerical Simulation and Multiple Surrogates , 2005 .

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

[6]  Federico Delfino,et al.  Data-Driven Photovoltaic Power Production Nowcasting and Forecasting for Polygeneration Microgrids , 2018, IEEE Systems Journal.

[7]  Yan Su,et al.  Analysis of daily solar power prediction with data-driven approaches , 2014 .

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

[9]  Yan Su,et al.  Forecasting the daily power output of a grid-connected photovoltaic system based on multivariate adaptive regression splines , 2016 .

[10]  Fabio Viola,et al.  Day-ahead forecasting for photovoltaic power using artificial neural networks ensembles , 2016, 2016 IEEE International Conference on Renewable Energy Research and Applications (ICRERA).

[11]  Xiaojuan Han,et al.  Day-ahead forecasting of photovoltaic output power with similar cloud space fusion based on incomplete historical data mining , 2017 .

[12]  Lei Wang,et al.  An ANN-based Approach for Forecasting the Power Output of Photovoltaic System , 2011 .

[13]  T. Takashima,et al.  Use of support vector regression and numerically predicted cloudiness to forecast power output of a photovoltaic power plant in Kitakyushu, Japan , 2012 .

[14]  Luca Delle Monache,et al.  Short-term photovoltaic power forecasting using Artificial Neural Networks and an Analog Ensemble , 2017 .

[15]  Tao Zhang,et al.  Predicating photovoltaic power generation using an improved hybrid heuristic method , 2016, 2016 Sixth International Conference on Information Science and Technology (ICIST).

[16]  Liuchen Chang,et al.  Short-term photovoltaic output forecasting model for economic dispatch of power system incorporating large-scale photovoltaic plant , 2013, 2013 IEEE Energy Conversion Congress and Exposition.

[17]  R. Haftka,et al.  Ensemble of surrogates , 2007 .

[18]  M. Rais-Rohani,et al.  Ensemble of metamodels with optimized weight factors , 2008 .

[19]  X. Wen,et al.  A wavelet-coupled support vector machine model for forecasting global incident solar radiation using limited meteorological dataset , 2016 .