Stability of implicit neural networks for long-term forecasting in dynamical systems

Abstract

Forecasting physical signals in long time range is among the most challenging tasks in Partial Differential Equations (PDEs) research. To circumvent limitations of traditional solvers, many different Deep Learning methods have been proposed. They are all based on auto-regressive methods and exhibit stability issues. Drawing inspiration from the stability property of implicit numerical schemes, we introduce a stable auto-regressive implicit neural network. We develop a theory based on the stability definition of schemes to ensure the stability in forecasting of this network. It leads us to introduce hard constraints on its weights and propagate the dynamics in the latent space. Our experimental results validate our stability property, and show improved results at long-term forecasting for two transports PDEs.

arXiv commentICLR 2023 Workshop on Physics for Machine Learning

References in corpus (9)

• Fourier Neural Operator for Parametric Partial Differential Equations
• Message Passing Neural PDE Solvers
• Learning Dynamical Systems from Partial Observations
• Comparing recurrent and convolutional neural networks for predicting wave propagation
• IMEXnet: A Forward Stable Deep Neural Network
• Physics-Embedded Neural Networks: Graph Neural PDE Solvers with Mixed Boundary Conditions
• Robust learning with implicit residual networks
• Simulating Surface Wave Dynamics with Convolutional Networks
• $\xi$-torch: differentiable scientific computing library