Gaussian process hydrodynamics
From MaRDI portal
Publication:6132295
Gaussian processes (60G15) Probabilistic methods, particle methods, etc. for boundary value problems involving PDEs (65N75) Navier-Stokes equations (35Q30) Navier-Stokes equations for incompressible viscous fluids (76D05) Nonlinear boundary value problems for ordinary differential equations (34B15) Spline approximation (41A15) Probabilistic methods, particle methods, etc. for initial value and initial-boundary value problems involving PDEs (65M75) Spectral, collocation and related methods for boundary value problems involving PDEs (65N35) Kernel operators (47B34)
Abstract: We present a Gaussian Process (GP) approach (Gaussian Process Hydrodynamics, GPH) for approximating the solution of the Euler and Navier-Stokes equations. As in Smoothed Particle Hydrodynamics (SPH), GPH is a Lagrangian particle-based approach involving the tracking of a finite number of particles transported by the flow. However, these particles do not represent mollified particles of matter but carry discrete/partial information about the continuous flow. Closure is achieved by placing a divergence-free GP prior on the velocity field and conditioning on vorticity at particle locations. Known physics (e.g., the Richardson cascade and velocity-increments power laws) is incorporated into the GP prior through physics-informed additive kernels. This approach allows us to coarse-grain turbulence in a statistical manner rather than a deterministic one. By enforcing incompressibility and fluid/structure boundary conditions through the selection of the kernel, GPH requires much fewer particles than SPH. Since GPH has a natural probabilistic interpretation, numerical results come with uncertainty estimates enabling their incorporation into a UQ pipeline and the adding/removing of particles (quantas of information) in an adapted manner. The proposed approach is amenable to analysis, it inherits the complexity of state-of-the-art solvers for dense kernel matrices, and it leads to a natural definition of turbulence as information loss. Numerical experiments support the importance of selecting physics-informed kernels and illustrate the major impact of such kernels on accuracy and stability. Since the proposed approach has a Bayesian interpretation, it naturally enables data assimilation and making predictions and estimations based on mixing simulation data with experimental data.
Recommendations
- Gaussian processes for computer experiments
- Gaussian processes and Gaussian random fields
- Numerical Gaussian processes for time-dependent and nonlinear partial differential equations
- A new class of high-order methods for fluid dynamics simulations using Gaussian process modeling: one-dimensional case
- Gaussian processes and mixed volumes
- Applying Bayesian optimization with Gaussian process regression to computational fluid dynamics problems
- Solving and learning nonlinear PDEs with Gaussian processes
- scientific article; zbMATH DE number 3850197
- Enhanced Gaussian processes and applications
Cites work
- scientific article; zbMATH DE number 5787960 (Why is no real title available?)
- scientific article; zbMATH DE number 3601500 (Why is no real title available?)
- Bayesian numerical homogenization
- Can a finite element method perform arbitrarily badly?
- Can the atmospheric kinetic energy spectrum be explained by two-dimensional turbulence?
- Compression, inversion, and approximate PCA of dense kernel matrices at near-linear computational complexity
- Computational graph completion
- Consistency of empirical Bayes and kernel flow for hierarchical parameter estimation
- De-noising by thresholding operator adapted wavelets
- Do ideas have shape? Idea registration as the continuous limit of artificial neural networks
- Gamblets for opening the complexity-bottleneck of implicit schemes for hyperbolic and parabolic ODEs/PDEs with rough coefficients
- Hidden fluid mechanics: Learning velocity and pressure fields from flow visualizations
- Kernel flows: from learning kernels from data into the abyss
- Kernel mode decomposition and the programming of kernels
- Kernels for vector-valued functions: a review
- Learning ``best kernels from data in Gaussian process regression. With application to aerodynamics
- Learning dynamical systems from data: a simple cross-validation perspective. I: Parametric kernel flows
- Learning dynamical systems from data: a simple cross-validation perspective. III: Irregularly-sampled time series
- Multigrid with Rough Coefficients and Multiresolution Operator Decomposition from Hierarchical Information Games
- One-shot learning of stochastic differential equations with data adapted kernels
- Operator-adapted wavelets, fast solvers, and numerical homogenization. From a game theoretic approach to numerical approximation and algorithm design
- Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
- Reproducing kernel Hilbert spaces associated with analytic translation-invariant Mercer kernels
- Scattered Data Approximation
- Sixth problem of the millennium: Navier-Stokes equations, existence and smoothness
- Smoothed particle hydrodynamics (SPH): an overview and recent developments
- Smoothed particle hydrodynamics: theory and application to non-spherical stars
- Solving and learning nonlinear PDEs with Gaussian processes
- Sparse Cholesky Factorization by Kullback--Leibler Minimization
- Stability of two-dimensional Navier-Stokes motions in the periodic case
- Statistical Numerical Approximation
- The anisotropic Lagrangian averaged Euler and Navier-Stokes equations
- Two-dimensional turbulence
- Vortex Methods
- Vortex methods for flow simulation
Cited in
(3)
This page was built for publication: Gaussian process hydrodynamics
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6132295)
