Physics-enhanced machine learning strategies for applied mechanics

This project will focus on addressing two fundamental challenges in physics-enhanced machine learning strategies in applied mechanics: (i) Overcoming poor generalisation performance and physically inconsistent or implausible predictions of machine learning models in applied mechanics by developing approaches integrating physics (first principles) knowledge through biases within Machine Learning (ML) algorithms to inform physics (e,g. identification of unknown constitutive laws and nonlinearities from measurements and physics-knowledge). (ii) Identification of incorrect prior physics assumption (e.g. wrong constitutive model) in the physics-enhanced machine learning algorithm.