Presentations

Abstract

It is essential to be able to properly simulate wake effects during the wind farm design in order to optimize the wind farm lay-out, and to estimate the expected annual energy yield. However, calibrating wake models presents significant challenges. In literature, different types of tuning procedures have been presented. These methodologies typically integrate scada data or LiDAR data to calibrate the model match the observations. This work aims to compare the results obtained from LiDAR-based and scada-based hyperparameter tuning frameworks.. This is illustrated on various analytical wake models.