In this section
@article{Peters:2027:10.1016/j.epsr.2026.113733,
author = {Peters, A and Pinson, P and Banovic, O and Bouchkati, S and Ulbig, A},
doi = {10.1016/j.epsr.2026.113733},
journal = {Electric Power Systems Research},
title = {Developing an open-source and Europe-wide wind power forecasting platform},
url = {http://dx.doi.org/10.1016/j.epsr.2026.113733},
volume = {263},
year = {2027}
}
TY - JOUR
AB - This work presents the novel open-source platform Wind Power Forecast for open and accessible wind power forecasting across Europe. The platform is motivated by the limited accessibility of wind power forecasting tools, which are often not openly available, difficult to adapt to arbitrary wind power plants (WPPs), or restricted by limited data availability. To address this, a model based on Multi-Layer Perceptrons (MLPs) is developed to generalise across arbitrary WPPs, enabling forecasts for existing and custom WPPs without asset-specific training. In this work, ”real-time forecast” refers to forecasts that are automatically updated using the most recent available weather forecast data. The model is trained on wind speeds at 100 m and WPP data, namely turbine type, hub height, age, and recorded production data, and it outperforms a power-curve benchmark. A web application renders the forecasts publicly available, visualises them on an interactive map, and allows users to define custom WPPs. In addition, a crowdsourcing feature enables users to upload production time series in order to extend the database and continuously improve the model. The platform is intended to particularly benefit smaller stakeholders and researchers by improving accessibility, transparency, interpretability, and opportunities for community-based contribution.
AU - Peters,A
AU - Pinson,P
AU - Banovic,O
AU - Bouchkati,S
AU - Ulbig,A
DO - 10.1016/j.epsr.2026.113733
PY - 2027///
SN - 0378-7796
TI - Developing an open-source and Europe-wide wind power forecasting platform
T2 - Electric Power Systems Research
UR - http://dx.doi.org/10.1016/j.epsr.2026.113733
VL - 263
ER -