Learning to Control PDEs with Differentiable Physics
- URL: http://arxiv.org/abs/2001.07457v1
- Date: Tue, 21 Jan 2020 11:58:41 GMT
- Title: Learning to Control PDEs with Differentiable Physics
- Authors: Philipp Holl, Vladlen Koltun, Nils Thuerey
- Abstract summary: We present a novel hierarchical predictor-corrector scheme which enables neural networks to learn to understand and control complex nonlinear physical systems over long time frames.
We demonstrate that our method successfully develops an understanding of complex physical systems and learns to control them for tasks involving PDEs.
- Score: 102.36050646250871
- License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
- Abstract: Predicting outcomes and planning interactions with the physical world are
long-standing goals for machine learning. A variety of such tasks involves
continuous physical systems, which can be described by partial differential
equations (PDEs) with many degrees of freedom. Existing methods that aim to
control the dynamics of such systems are typically limited to relatively short
time frames or a small number of interaction parameters. We present a novel
hierarchical predictor-corrector scheme which enables neural networks to learn
to understand and control complex nonlinear physical systems over long time
frames. We propose to split the problem into two distinct tasks: planning and
control. To this end, we introduce a predictor network that plans optimal
trajectories and a control network that infers the corresponding control
parameters. Both stages are trained end-to-end using a differentiable PDE
solver. We demonstrate that our method successfully develops an understanding
of complex physical systems and learns to control them for tasks involving PDEs
such as the incompressible Navier-Stokes equations.
Related papers
- Pretraining Codomain Attention Neural Operators for Solving Multiphysics PDEs [85.40198664108624]
We propose Codomain Attention Neural Operator (CoDA-NO) to solve multiphysics problems with PDEs.
CoDA-NO tokenizes functions along the codomain or channel space, enabling self-supervised learning or pretraining of multiple PDE systems.
We evaluate CoDA-NO's potential as a backbone for learning multiphysics PDEs over multiple systems by considering few-shot learning settings.
arXiv Detail & Related papers (2024-03-19T08:56:20Z) - iPINNs: Incremental learning for Physics-informed neural networks [66.4795381419701]
Physics-informed neural networks (PINNs) have recently become a powerful tool for solving partial differential equations (PDEs)
We propose incremental PINNs that can learn multiple tasks sequentially without additional parameters for new tasks and improve performance for every equation in the sequence.
Our approach learns multiple PDEs starting from the simplest one by creating its own subnetwork for each PDE and allowing each subnetwork to overlap with previously learnedworks.
arXiv Detail & Related papers (2023-04-10T20:19:20Z) - Mixed formulation of physics-informed neural networks for
thermo-mechanically coupled systems and heterogeneous domains [0.0]
Physics-informed neural networks (PINNs) are a new tool for solving boundary value problems.
Recent investigations have shown that when designing loss functions for many engineering problems, using first-order derivatives and combining equations from both strong and weak forms can lead to much better accuracy.
In this work, we propose applying the mixed formulation to solve multi-physical problems, specifically a stationary thermo-mechanically coupled system of equations.
arXiv Detail & Related papers (2023-02-09T21:56:59Z) - Semi-supervised Learning of Partial Differential Operators and Dynamical
Flows [68.77595310155365]
We present a novel method that combines a hyper-network solver with a Fourier Neural Operator architecture.
We test our method on various time evolution PDEs, including nonlinear fluid flows in one, two, and three spatial dimensions.
The results show that the new method improves the learning accuracy at the time point of supervision point, and is able to interpolate and the solutions to any intermediate time.
arXiv Detail & Related papers (2022-07-28T19:59:14Z) - Multi-resolution partial differential equations preserved learning
framework for spatiotemporal dynamics [11.981731023317945]
Physics-informed deep learning (PiDL) addresses these challenges by incorporating physical principles into the model.
We propose to leverage physics prior knowledge by baking'' the discretized governing equations into the neural network architecture.
This method, embedding discretized PDEs through convolutional residual networks in a multi-resolution setting, largely improves the generalizability and long-term prediction.
arXiv Detail & Related papers (2022-05-09T01:27:58Z) - Physics-constrained Unsupervised Learning of Partial Differential
Equations using Meshes [1.066048003460524]
Graph neural networks show promise in accurately representing irregularly meshed objects and learning their dynamics.
In this work, we represent meshes naturally as graphs, process these using Graph Networks, and formulate our physics-based loss to provide an unsupervised learning framework for partial differential equations (PDE)
Our framework will enable the application of PDE solvers in interactive settings, such as model-based control of soft-body deformations.
arXiv Detail & Related papers (2022-03-30T19:22:56Z) - Characterizing possible failure modes in physics-informed neural
networks [55.83255669840384]
Recent work in scientific machine learning has developed so-called physics-informed neural network (PINN) models.
We demonstrate that, while existing PINN methodologies can learn good models for relatively trivial problems, they can easily fail to learn relevant physical phenomena even for simple PDEs.
We show that these possible failure modes are not due to the lack of expressivity in the NN architecture, but that the PINN's setup makes the loss landscape very hard to optimize.
arXiv Detail & Related papers (2021-09-02T16:06:45Z) - Encoding physics to learn reaction-diffusion processes [18.187800601192787]
We show how a deep learning framework that encodes given physics structure can be applied to a variety of problems regarding the PDE system regimes.
The resultant learning paradigm that encodes physics shows high accuracy, robustness, interpretability and generalizability demonstrated via extensive numerical experiments.
arXiv Detail & Related papers (2021-06-09T03:02:20Z) - Controlling nonlinear dynamical systems into arbitrary states using
machine learning [77.34726150561087]
We propose a novel and fully data driven control scheme which relies on machine learning (ML)
Exploiting recently developed ML-based prediction capabilities of complex systems, we demonstrate that nonlinear systems can be forced to stay in arbitrary dynamical target states coming from any initial state.
Having this highly flexible control scheme with little demands on the amount of required data on hand, we briefly discuss possible applications that range from engineering to medicine.
arXiv Detail & Related papers (2021-02-23T16:58:26Z) - Neural-PDE: A RNN based neural network for solving time dependent PDEs [6.560798708375526]
Partial differential equations (PDEs) play a crucial role in studying a vast number of problems in science and engineering.
We propose a sequence deep learning framework called Neural-PDE, which allows to automatically learn governing rules of any time-dependent PDE system.
In our experiments the Neural-PDE can efficiently extract the dynamics within 20 epochs training, and produces accurate predictions.
arXiv Detail & Related papers (2020-09-08T15:46:00Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.