Abstract:This paper proposes the use of Artificial Neural Networks (ANNs), specifically Physics-Informed Neural Networks (PINNs), for dynamic surrogate modelling of wind turbines. PINNs offer the flexibility to model complex relationships while incorporating physics-based constraints, enabling accurate representation of wind turbine dynamics. In this paper, a PINN-based surrogate model is developed for the Blade Element Momentum (BEM) aerodynamic model used in state-of-the-art numerical wind turbine simulations. The PINN model replaces the time-consuming root-finding process in BEM with high-dimensional regression, significantly improving computational efficiency. The PINN model is trained using data generated from a numerical model of the IEA-15MW reference wind turbine, and its performance is compared against conventional data-driven Neural Network (DDNN) models. The proposed surrogate model provides more efficient and accurate evaluations of wind turbine responses compared with traditional surrogate modelling approaches. A significant computational advantage is obtained by using the developed surrogate models with a forty-fold speedup demonstrated compared to the BEM model. Replacing the BEM model with the PINN-based surrogate model for load computation in the numerical model used for dynamic analysis results in an overall reduction of 35% in computational time for a complete dynamic simulation. This is a substantial improvement in efficiency without sacrificing accuracy — the maximum Mean Absolute Error (MAE) values for the surrogate models are of the order of 10−2 , which shows that the surrogate models can predict the angle of attack at any blade node with a discrepancy of less than 0.5° . The surrogate models significantly reduce computational time while maintaining high accuracy, making them a promising approach for simulating wind turbine dynamics, especially in fields such as reliability analysis or fatigue estimation where many simulations are necessary.

A neural network surrogate model for the performance assessment of a vertical structure subjected to non-stationary, tornadic wind loads

Some Results on the Vulnerability Assessment of HAWTs Subjected to Wind and Seismic Actions

Bayesian Approach for Typhoon-Induced Fragility Analysis of Real Overhead Transmission Lines

A Data‐driven Combined Deterministic‐stochastic Subspace Identification Method for Condition Assessment of Roof Structures Subjected to Strong Winds

Performance and Risk to Light-Framed Wood Residential Buildings Subjected to Tornadoes

Modelling of aero-mechanical response of wind turbine blade with damages by computational fluid dynamics, finite element analysis and Bayesian network

A surrogate model approach for associating wind farm load variations with turbine failures

Effectiveness Analysis of Structural Anomaly Diagnosis Based on ANN Model

Estimating nonlinear wind-induced response of roof cable nets by aeroelastic experiments and ML modeling

Estimation of rotor and main bearing loads using artificial neural networks

T2FL: An Efficient Model for Wind Turbine Fatigue Damage Prediction for the Two-Turbine Case

An artificial neural network model for fatigue damage analysis of wide-band non-Gaussian random processes

Physics guided wavelet convolutional neural network for wind-induced vibration modeling with application to structural dynamic reliability analysis

Time-Domain Structural Damage and Loss Estimates for Wind Loads: Road to Resilience-Targeted and Smart Buildings Design

An adaptive surrogate model to structural reliability analysis using deep neural network

Numerical Simulation of Dynamic Response of a Long-Span Bridge to Assess Its Vulnerability to Non-Synoptic Wind

In-event surrogate model training for regional transmission tower-line system dynamic responses prediction in realistic hurricane wind fields

Performance-based wind engineering assessment of a telecommunication mast utilizing Bayesian extreme wind velocity model

Wind Turbine Damage Equivalent Load Assessment Using Gaussian Process Regression Combining Measurement and Synthetic Data

Physics-Informed Neural Network surrogate model for bypassing Blade Element Momentum theory in wind turbine aerodynamic load estimation

Extrapolating low-occurrence strong wind speeds at pedestrian levels using artificial neural networks trained by a single turbulent dataset