8 papers
Alternating Levenberg-Marquardt Training of Physics-Informed Neural Networks with Fourier-Enhanced Features
Yulun Wu, Matthieu Barreau, Miguel Aguiar +1
Physics-informed neural networks (PINNs) often fail to accurately resolve partial differential equations (PDEs) with high-frequency or multi-scale solutions, as well as strongly no…
Locally Stable Neural ODEs with Characterized Region of Attraction
Alice Harting, Karl Henrik Johansson, Sophie Tarbouriech +1
We propose a class of neural ODEs that universally approximates locally exponentially stable dynamics and the region of attraction from trajectory data. The model dynamics are cons…
Physics-Informed Detection of Friction Anomalies in Satellite Reaction Wheels
Alejandro Penacho Riveiros, Nicola Bastianello, Karl H. Johansson +1
As the number of satellites in orbit has increased exponentially in recent years, ensuring their correct functionality has started to require automated methods to decrease human wo…
RHYME-XT: A Neural Operator for Spatiotemporal Control Systems
Marijn Ruiter, Miguel Aguiar, Jake Rap +2
We propose RHYME-XT, an operator-learning framework for surrogate modeling of spatiotemporal control systems governed by input-affine nonlinear partial integro-differential equatio…
Iterative Training of Physics-Informed Neural Networks with Fourier-enhanced Features
Yulun Wu, Miguel Aguiar, Karl H. Johansson +1
Spectral bias, the tendency of neural networks to learn low-frequency features first, is a well-known issue with many training algorithms for physics-informed neural networks (PINN…
Closed-Loop Neural Operator-Based Observer of Traffic Density
Alice Harting, Karl Henrik Johansson, Matthieu Barreau
We consider the problem of traffic density estimation with sparse measurements from stationary roadside sensors. Our approach uses Fourier neural operators to learn macroscopic tra…