Impact of PV forecasts uncertainty in batteries management in microgrids

This paper is motivated by the question of the impact that uncertainty in PV forecasts has in forecast-based battery schedule optimisation in microgrids in presence of network constraints. We examine a specific case where forecast accuracy can be impacted by the lack of enough data history to finetune the forecasting models. This situation can be expected to be frequent with new PV installations. A probabilistic PV production forecast algorithm is used in combination with a battery schedule optimisation algorithm. The size of the learning dataset of the forecast algorithm is modified in order to simulate the application of the system to new plants and the impact on the performance in the management of the battery is analysed.

[1]  T. Gneiting.  Quantiles as optimal point forecasts , 2011 .

[2]  Su Shi,et al.  A practical model for single-step power prediction of grid-connected PV plant using artificial neural network , 2011, 2011 IEEE PES Innovative Smart Grid Technologies.

[3]  A. Raftery,et al.  Probabilistic forecasts, calibration and sharpness , 2007 .

[4]  Andrea Michiorri,et al.  THE IMPACT OF AVAILABLE DATA HISTORY ON THE PERFORMANCE OF PHOTOVOLTAIC GENERATION FORECASTING MODELS , 2013 .

[5]  Zuyi Li,et al.  A Fuzzy Optimization Approach to PV/Battery Scheduling with Uncertainty in PV Generation , 2006, 2006 38th North American Power Symposium.

[6]  Henrik Madsen,et al.  Online short-term solar power forecasting , 2009 .

[7]  M. Etezadi-Amoli,et al.  Practical approach for sub-hourly and hourly prediction of PV power output , 2010, North American Power Symposium 2010.

[8]  M. P. Moghaddam,et al.  Economic viability of NaS battery plant in a competitive electricity market , 2009, 2009 International Conference on Clean Electrical Power.

[9]  G. Ippoliti,et al.  Solar irradiation forecasting using RBF networks for PV systems with storage , 2012, 2012 IEEE International Conference on Industrial Technology.

[10]  G. Kariniotakis,et al.  A Stochastic Dynamic Programming Model for Optimal Use of Local Energy Resources in a Market Environment , 2007, 2007 IEEE Lausanne Power Tech.

[11]  Chen Changsong,et al.  Forecasting power output for grid-connected photovoltaic power system without using solar radiation measurement , 2010, The 2nd International Symposium on Power Electronics for Distributed Generation Systems.

[12]  R. Koenker.  Quantile Regression: Name Index , 2005 .

[13]  Pierre Pinson,et al.  Non‐parametric probabilistic forecasts of wind power: required properties and evaluation , 2007 .

[14]  Nicolai Meinshausen,et al.  Quantile Regression Forests , 2006, J. Mach. Learn. Res..

[15]  Jan Kloppenborg Møller,et al.  ARTICLE IN PRESS Computational Statistics & Data Analysis ( ) – Time-adaptive quantile regression , 2022 .