Comparison of Statistical-Based and Data-Driven-Based Scenario Generation of PV Power for Stochastic Day-Ahead Battery Scheduling

The day-ahead PV power generation scenarios, which represent the possible output of PV, have a significant impact on the scheduling of the available flexibility in a smart building. In this paper, a scenario generation approach for the day-ahead production of a single PV system is presented. The proposed method is important in the context of single buildings where the self-consumption has to be optimized. LASSO, which is a data-driven method, is used in order to select relevant quantiles to capture CDF. Moreover, a statistical-based scenario generation is applied in order to compare the performance of the proposed method. The generated scenarios for day-ahead PV generation are used in a stochastic problem to minimize the expected operational cost of a building and manage the flexibility, which is battery in this case study. Finally, the proposed method has been applied to a real PV installation on the rooftop of EnergyVille-1, a research institute. The simulation results demonstrate that the proposed method is able to capture the dynamics of the system even with low number of scenarios, which leads to reduce the computational time of stochastic problem.

[1]  Xi-Yuan Ma,et al.  Scenario Generation of Wind Power Based on Statistical Uncertainty and Variability , 2013, IEEE Transactions on Sustainable Energy.

[2]  A. G. Bakirtzis,et al.  Application of time series and artificial neural network models in short-term forecasting of PV power generation , 2013, 2013 48th International Universities' Power Engineering Conference (UPEC).

[3]  Chul-Hwan Kim,et al.  Application of neural network to 24-hour-ahead generating power forecasting for PV system , 2008, 2008 IEEE Power and Energy Society General Meeting - Conversion and Delivery of Electrical Energy in the 21st Century.

[4]  Olivier Deblecker,et al.  Optimal operation of an energy management system for a grid-connected smart building considering photovoltaics’ uncertainty and stochastic electric vehicles’ driving schedule , 2018 .

[5]  Kenneth Bruninx,et al.  Improved Modeling of Unit Commitment Decisions under Uncertainty , 2016 .

[6]  Geert Deconinck,et al.  Ensemble Machine Learning Forecaster for Day Ahead PV System Generation , 2019, 2019 IEEE 7th International Conference on Smart Energy Grid Engineering (SEGE).

[7]  Conversion and delivery of electrical energy in the 21st century , 2008, 2008 IEEE Power and Energy Society General Meeting - Conversion and Delivery of Electrical Energy in the 21st Century.

[8]  Wencong Su,et al.  Stochastic Energy Scheduling in Microgrids With Intermittent Renewable Energy Resources , 2014, IEEE Transactions on Smart Grid.

[9]  David L. Woodruff,et al.  Constructing probabilistic scenarios for wide-area solar power generation , 2018 .

[10]  G. G. Stokes.  "J." , 1890, The New Yale Book of Quotations.

[11]  H. Madsen,et al.  From probabilistic forecasts to statistical scenarios of short-term wind power production , 2009 .

[12]  P. Pinson,et al.  Generation and evaluation of space–time trajectories of photovoltaic power , 2016, 1603.06649.

[13]  Hedayat Saboori,et al.  Stochastic optimal battery storage sizing and scheduling in home energy management systems equipped with solar photovoltaic panels , 2017 .

[14]  Michal Kaut,et al.  Evaluation of scenario-generation methods for stochastic programming , 2007 .

[15]  Johan A. K. Suykens,et al.  Spatio-temporal feature selection for black-box weather forecasting , 2016, ESANN.

[16]  Alexander Shapiro,et al.  Lectures on Stochastic Programming: Modeling and Theory , 2009 .

[17]  Johan Driesen,et al.  A 368-kWp Grid-Connected PV System: Known and Hidden Losses , 2017 .