Wind–Wave Environmental Conditions and Structural Response Dataset for Offshore Wind Turbine Surrogate Modeling

This dataset provides environmental input parameters and corresponding structural responses of a monopile-supported NREL 5 MW offshore wind turbine under tropical cyclone (TC) and normal operating conditions. The dataset is intended to support the development, training, and validation of surrogate models for structural response prediction and reliability assessment, considering both ultimate limit state (ULS) and fatigue limit state (FLS) performance. Environmental inputs Each input file contains six environmental variables: Wind speed at 90 m hub height (m/s) Wind direction (°) Significant wave height (m) Spectral peak wave period (s) Wave direction (°) Structural response outputs Two categories of structural responses are provided: Tropical cyclone (TC) conditions: Maximum tower-base bending moment responses. These responses are used for extreme-load assessment under ULS conditions. Normal operating conditions: Damage equivalent loads (DELs) at 0°. These quantities support fatigue assessment under FLS conditions. Dataset organization The dataset contains separate training and validation files for TC and normal conditions: Condition Training files Test files TC TCtrain_input.csv, TCtrain_output.csv TCvali_input.csv, TCvali_output.csv Normal Normaltraning_input.csv, Normaltraning_output.csv Normalvali_input.csv, Normalvali_output.csv The training datasets contain 9,330 paired input–output records for TC conditions and 9,386 paired records for normal conditions. The test datasets comprise 38 TC environmental conditions and 46 normal environmental conditions, each associated with 150 structural response realizations generated using different random seeds. The test data are organized by environmental condition, with all 150 realizations for each condition arranged consecutively. Specifically, records 1–150 correspond to the first environmental input, records 151–300 to the second, and so forth. These data facilitate the development and evaluation of deterministic and stochastic surrogate models, including the characterization of structural response variability under identical environmental conditions.