Lagrangian Neural Networks

Abstract

Accurate models of the world are built upon notions of its underlying symmetries. In physics, these symmetries correspond to conservation laws, such as for energy and momentum. Yet even though neural network models see increasing use in the physical sciences, they struggle to learn these symmetries. In this paper, we propose Lagrangian Neural Networks (LNNs), which can parameterize arbitrary Lagrangians using neural networks. In contrast to models that learn Hamiltonians, LNNs do not require canonical coordinates, and thus perform well in situations where canonical momenta are unknown or difficult to compute. Unlike previous approaches, our method does not restrict the functional form of learned energies and will produce energy-conserving models for a variety of tasks. We test our approach on a double pendulum and a relativistic particle, demonstrating energy conservation where a baseline approach incurs dissipation and modeling relativity without canonical coordinates where a Hamiltonian approach fails. Finally, we show how this model can be applied to graphs and continuous systems using a Lagrangian Graph Network, and demonstrate it on the 1D wave equation.

7 pages (+2 appendix). Published in ICLR 2020 Deep Differential Equations Workshop. Code at github.com/MilesCranmer/lagrangian_nns

References in corpus (10)

• Sequence to Sequence Learning with Neural Networks
• Fast and Accurate Modeling of Molecular Atomization Energies with Machine Learning
• Quantum-Chemical Insights from Deep Tensor Neural Networks
• Interaction Networks for Learning about Objects, Relations and Physics
• Hamiltonian Neural Networks
• Discovering Symbolic Models from Deep Learning with Inductive Biases
• Deep Lagrangian Networks: Using Physics as Model Prior for Deep Learning
• Hamiltonian Graph Networks with ODE Integrators
• Symplectic Recurrent Neural Networks
• A General Framework for Structured Learning of Mechanical Systems

Cited by in corpus (42)

• Solver-in-the-Loop: Learning from Differentiable Physics to Interact with Iterative PDE-Solvers
• AI Poincaré: Machine Learning Conservation Laws from Trajectories
• Characterizing possible failure modes in physics-informed neural networks
• Physics-constrained, low-dimensional models for MHD: First-principles and data-driven approaches
• Interpretable Sequence Learning for COVID-19 Forecasting
• Parameterized Neural Ordinary Differential Equations: Applications to Computational Physics Problems
• Symplectic Neural Networks in Taylor Series Form for Hamiltonian Systems
• Port-Hamiltonian Neural Networks for Learning Explicit Time-Dependent Dynamical Systems
• Lipschitz Recurrent Neural Networks
• Adaptable Hamiltonian neural networks
• Machine-Learning Non-Conservative Dynamics for New-Physics Detection
• Meta-Learning Symmetries by Reparameterization
• Image Inversion and Uncertainty Quantification for Constitutive Laws of Pattern Formation
• Selecting Data Augmentation for Simulating Interventions
• Hypersolvers: Toward Fast Continuous-Depth Models
• Linearly Constrained Neural Networks
• LagNetViP: A Lagrangian Neural Network for Video Prediction
• Deep Learning of Conjugate Mappings
• Beltrami Flow and Neural Diffusion on Graphs
• TorchDyn: A Neural Differential Equations Library
• Augmenting Differentiable Simulators with Neural Networks to Close the Sim2Real Gap
• Structure Preservation for the Deep Neural Network Multigrid Solver
• Physics-Integrated Variational Autoencoders for Robust and Interpretable Generative Modeling
• Neural Networks with Physics-Informed Architectures and Constraints for Dynamical Systems Modeling
• Sparsely constrained neural networks for model discovery of PDEs
• One-Shot Transfer Learning of Physics-Informed Neural Networks
• Learning Potentials of Quantum Systems using Deep Neural Networks
• Neural Vortex Method: from Finite Lagrangian Particles to Infinite Dimensional Eulerian Dynamics
• Climate Modeling with Neural Diffusion Equations
• MC-LSTM: Mass-Conserving LSTM
• The Autodidactic Universe
• Physics-enhanced Neural Networks in the Small Data Regime
• Learning Dynamics Models with Stable Invariant Sets
• Model discovery in the sparse sampling regime
• Neural Symplectic Integrator with Hamiltonian Inductive Bias for the Gravitational $N$-body Problem
• Analysis of ODE2VAE with Examples
• Physics-Augmented Learning: A New Paradigm Beyond Physics-Informed Learning
• Extracting Dynamical Models from Data
• Learning Canonical Transformations
• Nonseparable Symplectic Neural Networks
• Training Structured Mechanical Models by Minimizing Discrete Euler-Lagrange Residual
• Incorporating NODE with Pre-trained Neural Differential Operator for Learning Dynamics