Posters

Abstract

Wind power forecasting is crucial for the efficient management of wind farms and their integration into the power grid. Accurate forecasting can assist in streamlining their operations, reduce costs, and increase revenue. In this work, we present a novel approach to wind power forecasting that combines Weather Research & Forecasting (WRF) simulations with the Python PyWake tool. We perform WRF simulations on a daily basis covering different regions of the world to generate an operational high-resolution 10-day forecast. This meteorological data, which includes wind speed, wind direction, turbulence intensity, among others, is utilized as input to the PyWake model. The model calculates wake interactions between turbines in a wind farm using engineering models, in a faster and more cost-effective than Computational Fluid Dynamics (CFD) models. The integration of PyWake into the forecasting process results in enhanced precision in wind power forecasts. Having analyzed the effectiveness of our technique using actual data gathered from various locations worldwide, we can affirm that our method portrays the situation within the wind farm more accurately. The process is fully automated and provides daily updated power predictions, both on a turbine and windfarm level, via a web interface. Due to this fact, it is well-suited for applications in wind farm management and grid integration. In the future, we plan to take alerts on scheduled maintenances into account to offer more accurate predictions.