Deep Lagrangian Networks: Using Physics as Model Prior for Deep Learning

Abstract

Deep learning has achieved astonishing results on many tasks with large amounts of data and generalization within the proximity of training data. For many important real-world applications, these requirements are unfeasible and additional prior knowledge on the task domain is required to overcome the resulting problems. In particular, learning physics models for model-based control requires robust extrapolation from fewer samples - often collected online in real-time - and model errors may lead to drastic damages of the system. Directly incorporating physical insight has enabled us to obtain a novel deep model learning approach that extrapolates well while requiring fewer samples. As a first example, we propose Deep Lagrangian Networks (DeLaN) as a deep network structure upon which Lagrangian Mechanics have been imposed. DeLaN can learn the equations of motion of a mechanical system (i.e., system dynamics) with a deep network efficiently while ensuring physical plausibility. The resulting DeLaN network performs very well at robot tracking control. The proposed method did not only outperform previous model learning approaches at learning speed but exhibits substantially improved and more robust extrapolation to novel trajectories and learns online in real-time

Published at ICLR 2019

References in corpus (5)

• DGM: A deep learning algorithm for solving partial differential equations
• Hidden Physics Models: Machine Learning of Nonlinear Partial Differential Equations
• Physics Informed Deep Learning (Part I): Data-driven Solutions of Nonlinear Partial Differential Equations
• Graph networks as learnable physics engines for inference and control
• PDE-Net: Learning PDEs from Data

Cited by in corpus (26)

• Informed Machine Learning -- A Taxonomy and Survey of Integrating Knowledge into Learning Systems
• Continuous-time system identification with neural networks: Model structures and fitting criteria
• Modeling System Dynamics with Physics-Informed Neural Networks Based on Lagrangian Mechanics
• Lagrangian Neural Networks
• Interpretable Sequence Learning for COVID-19 Forecasting
• Port-Hamiltonian Neural Networks for Learning Explicit Time-Dependent Dynamical Systems
• Adaptable Hamiltonian neural networks
• Wasserstein Robust Reinforcement Learning
• Machine-Learning Non-Conservative Dynamics for New-Physics Detection
• Simplifying Hamiltonian and Lagrangian Neural Networks via Explicit Constraints
• Deep Hamiltonian networks based on symplectic integrators
• Deep Learning in Protein Structural Modeling and Design
• TorchDyn: A Neural Differential Equations Library
• Augmenting Differentiable Simulators with Neural Networks to Close the Sim2Real Gap
• Pontryagin Differentiable Programming: An End-to-End Learning and Control Framework
• Neural Networks with Physics-Informed Architectures and Constraints for Dynamical Systems Modeling
• Noisy Recurrent Neural Networks
• Learning Potentials of Quantum Systems using Deep Neural Networks
• Climate Modeling with Neural Diffusion Equations
• Controlling Neural Networks with Rule Representations
• Physics-enhanced Neural Networks in the Small Data Regime
• Physics-aware, probabilistic model order reduction with guaranteed stability
• Learning Large-Time-Step Molecular Dynamics with Graph Neural Networks
• Learning Principle of Least Action with Reinforcement Learning
• Nonseparable Symplectic Neural Networks
• Plug and Play, Model-Based Reinforcement Learning