Dataset accompanying the publication "Gaussian process surrogate modeling for efficient controller tuning and fatigue load prediction of the helix wake-mixing method"
Dataset accompanying the publication "Gaussian process surrogate modeling for efficient controller tuning and fatigue load prediction of the helix wake-mixing method"
Dataset and accompanying scripts associated with the publication ''Gaussian process surrogate modeling for efficient controller tuning and fatigue load prediction of the helix wake-mixing method'' by van der Hoek et al. (2026) in Wind Energy Science. There are two main scripts: 1) GPHelixPowerGain.m - This file loads a preprocessed dataset of turbine power measurements from different LES simulations. Each simulation was run with different settings (frequency and amplitude) of the Helix method. Using the dataset, multiple Gaussian process models are generated that represent the turbine powers as a function of the Strouhal number and pitch amplitude of the Helix method (see Section 3 of the paper). 2) GPLoadSurrogateExample.m - This file uses a precomputed loads database for the IEA22 MW turbine (LoadSurrogateDataset_IEA22MW_TUD.xlsx) to create a Gaussian process load surrogate model that can estimate turbine DELs based on inflow data (Section 4 of the paper). Several sets of hyperparameters have already been optimized and are stored in gp_hyperparameters.zip.