Aerodynamic force reconstruction using physics-informed Gaussian processes
Researchers have developed a probabilistic physics-informed machine learning method to reconstruct aerodynamic loads from noisy structural response data. This approach, demonstrated on the Great Belt East Bridge, avoids overfitting and the need for regularization. The technique shows strong agreement in predicting load magnitudes, phase angles, and peak values, offering broad applicability for modeling validation and future load estimation. AI
IMPACT Introduces a novel physics-informed machine learning approach for aerodynamic load reconstruction, potentially improving structural analysis and prognosis.