What is it about?
Spatio-temporal solar forecasting models can anticipate incoming clouds based on data from upwind sites. For locations with predominant wind patterns, the high correlations with upwind sites is so strong, that it is trivial to build effective models. However, despite the overall good performance, these models are unable to adapt whenever the non-predominant winds occur. Thus, this work proposes a framework as to integrate cloud-level wind data and specialized models are built for different specific winds.
Featured Image
Why is it important?
Integrating wind data should have a low overall impact for locations with predominant winds (since the instances the model is improved are reduced considered in, for example, a whole year). However, it can work as an enabler for spatio-temporal approaches in locations with more even wind patterns (where the correlations in the data is more diluted).
Perspectives
Despite the interesting results, it would be important to reassess the proposed framework in two additional contexts: i) in a location with more even wind patterns, where far greater accuracy improvements would be expected; and ii) in a data set where the sites have different tilt and orientation angles, as it would better represent the reality of photovoltaic systems in urban areas.
Rodrigo Amaro e Silva
Faculty of Sciences, University of Lisbon
Read the Original
This page is a summary of: A regime-based approach for integrating wind information in spatio-temporal solar forecasting models, Journal of Renewable and Sustainable Energy, September 2019, American Institute of Physics,
DOI: 10.1063/1.5098763.
You can read the full text:
Contributors
The following have contributed to this page
