Data-driven identification of latent port-Hamiltonian systems
Not on the map. Why?
Abstract (arXiv)
Conventional physics-based modeling techniques involve high effort, e.g., time and expert knowledge, while data-driven methods often lack interpretability, structure, and sometimes reliability. To mitigate this, we present a data-driven system identification framework that derives models in the port-Hamiltonian (pH) formulation. This formulation is suitable for multi-physical systems while guaranteeing the useful system theoretical properties of passivity and stability. Our framework combines linear and nonlinear reduction with structured, physics-motivated system identification. In this process, high-dimensional state data obtained from possibly nonlinear systems serves as input for an autoencoder, which then performs two tasks: (i) nonlinearly transforming and (ii) reducing this data onto a low-dimensional latent space. In this space, a linear pH system, that satisfies the pH properties per construction, is parameterized by the weights of a neural network. The mathematical requirements are met by defining the pH matrices through Cholesky factorizations. The neural networks that define the coordinate transformation and the pH system are identified in a joint optimization process to match the dynamics observed in the data while defining a linear pH system in the latent space. The learned, low-dimensional pH system can describe even nonlinear systems and is rapidly computable due to its small size. The method is exemplified by a parametric mass-spring-damper and a nonlinear pendulum example, as well as the high-dimensional model of a disc brake with linear thermoelastic behavior.
Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 40 of its 97 references that are in the map. First posted on arXiv 15 Aug 2024.
Builds on
40 of 97 references.
References · 40
- 2014Adam: A Method for Stochastic Optimizationcited by 36,960
- 2017Deep learning for universal linear embeddings of nonlinear dynamicscited by 491
- 2021Modern Koopman Theory for Dynamical Systemscited by 269
- 2017Hidden Physics Models: Machine Learning of Nonlinear Partial Differential Equationscited by 208
- 2020Physics-informed learning of governing equations from scarce datacited by 167
- 2020A comprehensive deep learning-based approach to reduced order modeling of nonlinear time-dependent parametrized PDEscited by 105
- 2020A fast and accurate physics-informed neural network reduced order model with shallow masked autoencodercited by 80
- 2021POD-DL-ROM: enhancing deep learning-based reduced order models for nonlinear parametrized PDEs by proper orthogonal decompositioncited by 79
- 2021Port-Hamiltonian Neural Networks for Learning Explicit Time-Dependent Dynamical Systemscited by 66
- 2022Discovering Governing Equations from Partial Measurements with Deep Delay Autoencoderscited by 58
- 2023A graph convolutional autoencoder approach to model order reduction for parametrized PDEscited by 44
- 2022Reduced order modeling of parametrized systems through autoencoders and SINDy approach: continuation of periodic solutionscited by 29
28 more
- 2021A Comparison of Neural Network Architectures for Data-Driven Reduced-Order Modelingcited by 22
- 2023Machine Learning for Partial Differential Equationscited by 20
- 2022Pseudo-Hamiltonian Neural Networks with State-Dependent External Forcescited by 17
- 2022Compositional Learning of Dynamical System Models Using Port-Hamiltonian Neural Networkscited by 17
- 2023Learning Nonlinear Projections for Reduced-Order Modeling of Dynamical Systems using Constrained Autoencoderscited by 16
- 2023Gaussian Process Port-Hamiltonian Systems: Bayesian Learning with Physics Priorcited by 15
- 2023Symplectic model reduction of Hamiltonian systems using data-driven quadratic manifoldscited by 13
- 2023Guaranteed Stable Quadratic Models and their applications in SINDy and Operator Inferencecited by 10
- 2022Port-Hamiltonian Dynamic Mode Decompositioncited by 9
- 2024VENI, VINDy, VICI: a generative reduced-order modeling framework with uncertainty quantificationcited by 8
- 2023Neural Autoencoder-Based Structure-Preserving Model Order Reduction and Control Design for High-Dimensional Physical Systemscited by 7
- 2016On computing the distance to stability for matrices using linear dissipative Hamiltonian systemscited by 6
- 2023Low-dimensional Data-based Surrogate Model of a Continuum-mechanical Musculoskeletal System Based on Non-intrusive Model Order Reductioncited by 6
- 2023Data-Driven Identification of Quadratic Representations for Nonlinear Hamiltonian Systems using Weakly Symplectic Liftingscited by 6
- 2020A Non-Intrusive Method to Inferring Linear Port-Hamiltonian Realizations using Time-Domain Datacited by 5
- 2023Symplectic Autoencoders for Model Reduction of Hamiltonian Systemscited by 5
- 2018Robust port-Hamiltonian representations of passive systemscited by 4
- 2020SOBMOR: Structured Optimization-Based Model Order Reductioncited by 4
- 2021Port-Hamiltonian System Identification from Noisy Frequency Response Datacited by 3
- 2018Discrete-time port-Hamiltonian systems: A definition based on symplectic integrationcited by 2
- 2020Error bounds for port-Hamiltonian model and controller reduction based on system balancingcited by 2
- 2024Data-driven identification of port-Hamiltonian DAE systems by Gaussian processescited by 2
- 2017Computing nearest stable matrix pairscited by 1
- 2019Model reduction techniques for linear constant coefficient port-Hamiltonian differential-algebraic systemscited by 1
- 2019Structure-preserving Interpolatory Model Reduction for Port-Hamiltonian Differential-Algebraic Systemscited by 1
- 2020Extended balancing of continuous LTI systems: a structure-preserving approachcited by 1
- 2022Learning port-Hamiltonian systems -- algorithmscited by 1
- 2024Multi-Hierarchical Surrogate Learning for Structural Dynamical Crash Simulations Using Graph Convolutional Neural Networkscited by 1
Built on by
Citing papers per year: 2025: 1; 2026: 2.
3 papers.