Nonlinear Dynamic
Nonlinear dynamics research focuses on understanding and modeling systems whose behavior doesn't scale linearly with input changes, aiming to predict and control their complex evolution. Current efforts concentrate on developing data-driven methods, employing techniques like sparse identification of nonlinear dynamics (SINDy), physics-informed neural networks (PINNs), and echo state networks (ESNs), often enhanced with Bayesian inference or active learning strategies to improve accuracy and efficiency, especially in the presence of noise and limited data. These advancements have significant implications for diverse fields, enabling more accurate modeling and control of systems ranging from mechanical structures and fluid flows to biological processes and AI agents.
Papers
Physics-Informed Neural Networks for Time-Domain Simulations: Accuracy, Computational Cost, and Flexibility
Jochen Stiasny, Spyros Chatzivasileiadis
Real-Time Measurement-Driven Reinforcement Learning Control Approach for Uncertain Nonlinear Systems
Mohammed Abouheaf, Derek Boase, Wail Gueaieb, Davide Spinello, Salah Al-Sharhan