-
Spectral measures and iterative bounds for effective diffusivity of steady and space-time periodic flows
Authors:
N. B. Murphy,
D. Hallman,
E. Cherkaev,
J. Xin,
K. M. Golden
Abstract:
Over three decades ago the advection-diffusion equation for a steady fluid velocity field was homogenized, leading to a Stieltjes integral representation for the effective diffusivity, which is given in terms of a spectral measure of a compact self-adjoint operator and the Péclet number of the fluid flow. This result was recently extended to space-time periodic flows, instead involving an unbounde…
▽ More
Over three decades ago the advection-diffusion equation for a steady fluid velocity field was homogenized, leading to a Stieltjes integral representation for the effective diffusivity, which is given in terms of a spectral measure of a compact self-adjoint operator and the Péclet number of the fluid flow. This result was recently extended to space-time periodic flows, instead involving an unbounded self-adjoint operator. Padé approximants provide rigorous upper and lower bounds for Stieltjes functions in terms of the moments of the spectral measure. However, with the lack of a method for calculating the moments of the spectral measure for general fluid velocity fields, the utility of this powerful mathematical framework for calculating bounds for the effective diffusivity has not been fully realized. Here we significantly expand the applicability of this framework by providing an iterative method that enables an arbitrary number of moments, hence bounds, to be calculated analytically in closed form for both spatially and space-time periodic flows. The method is demonstrated for periodic flows in two spatial dimensions. The known asymptotic behavior of the effective diffusivity for a steady flow is accurately captured by high order upper and lower bounds, demonstrating the ability of the method to provide accurate estimates for the effective diffusivity for a broad range of parameter values.
△ Less
Submitted 29 April, 2024;
originally announced April 2024.
-
Regularized Reduced Order Lippman-Schwinger-Lanczos Method for Inverse Scattering Problems in the Frequency Domain
Authors:
Justin Baker,
Elena Cherkaev,
Vladimir Druskin,
Shari Moskow,
Mikhail Zaslavsky
Abstract:
Inverse scattering has a broad applicability in quantum mechanics, remote sensing, geophysical, and medical imaging. This paper presents a robust direct reduced order model (ROM) method for solving inverse scattering problems based on an efficient approximation of the resolvent operator regularizing the Lippmann-Schwinger-Lanczos (LSL) algorithm. We show that the efficiency of the method relies up…
▽ More
Inverse scattering has a broad applicability in quantum mechanics, remote sensing, geophysical, and medical imaging. This paper presents a robust direct reduced order model (ROM) method for solving inverse scattering problems based on an efficient approximation of the resolvent operator regularizing the Lippmann-Schwinger-Lanczos (LSL) algorithm. We show that the efficiency of the method relies upon the weak dependence of the orthogonalized basis on the unknown potential in the Schrödinger equation by demonstrating that the Lanczos orthogonalization is equivalent to performing Gram-Schmidt on the ROM time snapshots. We then develop the LSL algorithm in the frequency domain with two levels of regularization. We show that the same procedure can be extended beyond the Schrödinger formulation to the Helmholtz equation, e.g., to imaging the conductivity using diffusive electromagnetic fields in conductive media with localized positive conductivity perturbations. Numerical experiments for Helmholtz and Schrödinger problems show that the proposed bi-level regularization scheme significantly improves the performance of the LSL algorithm, allowing for good reconstructions with noisy data and large data sets.
△ Less
Submitted 27 November, 2023;
originally announced November 2023.
-
Solving inverse scattering problems via reduced-order model embedding procedures
Authors:
Jörn Zimmerling,
Vladimir Druskin,
Murthy Guddati,
Elena Cherkaev,
Rob Remis
Abstract:
We present a reduced-order model (ROM) methodology for inverse scattering problems in which the reduced-order models are data-driven, i.e. they are constructed directly from data gathered by sensors. Moreover, the entries of the ROM contain localised information about the coefficients of the wave equation.
We solve the inverse problem by embedding the ROM in physical space. Such an approach is a…
▽ More
We present a reduced-order model (ROM) methodology for inverse scattering problems in which the reduced-order models are data-driven, i.e. they are constructed directly from data gathered by sensors. Moreover, the entries of the ROM contain localised information about the coefficients of the wave equation.
We solve the inverse problem by embedding the ROM in physical space. Such an approach is also followed in the theory of ``optimal grids,'' where the ROMs are interpreted as two-point finite-difference discretisations of an underlying set of equations of a first-order continuous system on this special grid. Here, we extend this line of work to wave equations and introduce a new embedding technique, which we call Krein embedding, since it is inspired by Krein's seminal work on vibrations of a string. In this embedding approach, an adaptive grid and a set of medium parameters can be directly extracted from a ROM and we show that several limitations of optimal grid embeddings can be avoided. Furthermore, we show how Krein embedding is connected to classical optimal grid embedding and that convergence results for optimal grids can be extended to this novel embedding approach. Finally, we also briefly discuss Krein embedding for open domains, that is, semi-infinite domains that extend to infinity in one direction.
△ Less
Submitted 14 June, 2023;
originally announced June 2023.
-
A generalized expansion method for computing Laplace-Beltrami eigenfunctions on manifolds
Authors:
Jackson C. Turner,
Elena Cherkaev,
Dong Wang
Abstract:
Eigendecomposition of the Laplace-Beltrami operator is instrumental for a variety of applications from physics to data science. We develop a numerical method of computation of the eigenvalues and eigenfunctions of the Laplace-Beltrami operator on a smooth bounded domain based on the relaxation to the Schrödinger operator with finite potential on a Riemannian manifold and projection in a special ba…
▽ More
Eigendecomposition of the Laplace-Beltrami operator is instrumental for a variety of applications from physics to data science. We develop a numerical method of computation of the eigenvalues and eigenfunctions of the Laplace-Beltrami operator on a smooth bounded domain based on the relaxation to the Schrödinger operator with finite potential on a Riemannian manifold and projection in a special basis. We prove spectral exactness of the method and provide examples of calculated results and applications, particularly, in quantum billiards on manifolds.
△ Less
Submitted 19 October, 2022;
originally announced October 2022.
-
A two-stage method for reconstruction of parameters in diffusion equations
Authors:
Xuesong Bai,
Elena Cherkaev,
Dong Wang
Abstract:
Parameter reconstruction for diffusion equations has a wide range of applications. In this paper, we proposed a two-stage scheme to efficiently solve conductivity reconstruction problems for steady-state diffusion equations with solution data measured inside the domain. The first stage is based on total variation regularization of the log diffusivity and the split Bregman iteration method. In the…
▽ More
Parameter reconstruction for diffusion equations has a wide range of applications. In this paper, we proposed a two-stage scheme to efficiently solve conductivity reconstruction problems for steady-state diffusion equations with solution data measured inside the domain. The first stage is based on total variation regularization of the log diffusivity and the split Bregman iteration method. In the second stage, we apply the K-means clustering for the reconstruction of ``blocky'' conductivity functions. The convergence of the scheme is theoretically proved and extensive numerical examples are shown to demonstrate the performance of the scheme.
△ Less
Submitted 25 June, 2022;
originally announced June 2022.
-
Two-scale cut-and-projection convergence for quasiperiodic monotone operators
Authors:
Niklas Wellander,
Sebastien Guenneau,
Elena Cherkaev
Abstract:
Averaging certain class of quasiperiodic monotone operators can be simplified to the periodic homogenization setting by map** the original quasiperiodic structure onto a periodic structure in a higher dimensional space using cut-and projection method. We characterize cut-and-projection convergence limit of the nonlinear monotone partial differential operator…
▽ More
Averaging certain class of quasiperiodic monotone operators can be simplified to the periodic homogenization setting by map** the original quasiperiodic structure onto a periodic structure in a higher dimensional space using cut-and projection method. We characterize cut-and-projection convergence limit of the nonlinear monotone partial differential operator $-\mathrm{div} \; σ\left({\bf x},\frac{{\bf R}{\bf x}}η, \nabla u_η\right)$ for a bounded sequence $u_η$ in $W^{1,p}_0(Ω)$, where $1<p < \infty$, $Ω$ is a bounded open subset in $R^n$ with Lipschitz boundary. We identify the homogenized problem with a local equation defined on the hyperplane in the higher-dimensional space. A new corrector result is established.
△ Less
Submitted 8 June, 2022;
originally announced June 2022.
-
Proximal Implicit ODE Solvers for Accelerating Learning Neural ODEs
Authors:
Justin Baker,
Hedi Xia,
Yiwei Wang,
Elena Cherkaev,
Akil Narayan,
Long Chen,
Jack Xin,
Andrea L. Bertozzi,
Stanley J. Osher,
Bao Wang
Abstract:
Learning neural ODEs often requires solving very stiff ODE systems, primarily using explicit adaptive step size ODE solvers. These solvers are computationally expensive, requiring the use of tiny step sizes for numerical stability and accuracy guarantees. This paper considers learning neural ODEs using implicit ODE solvers of different orders leveraging proximal operators. The proximal implicit so…
▽ More
Learning neural ODEs often requires solving very stiff ODE systems, primarily using explicit adaptive step size ODE solvers. These solvers are computationally expensive, requiring the use of tiny step sizes for numerical stability and accuracy guarantees. This paper considers learning neural ODEs using implicit ODE solvers of different orders leveraging proximal operators. The proximal implicit solver consists of inner-outer iterations: the inner iterations approximate each implicit update step using a fast optimization algorithm, and the outer iterations solve the ODE system over time. The proximal implicit ODE solver guarantees superiority over explicit solvers in numerical stability and computational efficiency. We validate the advantages of proximal implicit solvers over existing popular neural ODE solvers on various challenging benchmark tasks, including learning continuous-depth graph neural networks and continuous normalizing flows.
△ Less
Submitted 18 April, 2022;
originally announced April 2022.
-
Learning POD of Complex Dynamics Using Heavy-ball Neural ODEs
Authors:
Justin Baker,
Elena Cherkaev,
Akil Narayan,
Bao Wang
Abstract:
Proper orthogonal decomposition (POD) allows reduced-order modeling of complex dynamical systems at a substantial level, while maintaining a high degree of accuracy in modeling the underlying dynamical systems. Advances in machine learning algorithms enable learning POD-based dynamics from data and making accurate and fast predictions of dynamical systems. In this paper, we leverage the recently p…
▽ More
Proper orthogonal decomposition (POD) allows reduced-order modeling of complex dynamical systems at a substantial level, while maintaining a high degree of accuracy in modeling the underlying dynamical systems. Advances in machine learning algorithms enable learning POD-based dynamics from data and making accurate and fast predictions of dynamical systems. In this paper, we leverage the recently proposed heavy-ball neural ODEs (HBNODEs) [Xia et al. NeurIPS, 2021] for learning data-driven reduced-order models (ROMs) in the POD context, in particular, for learning dynamics of time-varying coefficients generated by the POD analysis on training snapshots generated from solving full order models. HBNODE enjoys several practical advantages for learning POD-based ROMs with theoretical guarantees, including 1) HBNODE can learn long-term dependencies effectively from sequential observations and 2) HBNODE is computationally efficient in both training and testing. We compare HBNODE with other popular ROMs on several complex dynamical systems, including the von Kármán Street flow, the Kurganov-Petrova-Popov equation, and the one-dimensional Euler equations for fluids modeling.
△ Less
Submitted 9 March, 2023; v1 submitted 24 February, 2022;
originally announced February 2022.
-
Wave-driven assembly of quasi periodic patterns of particles
Authors:
Elena Cherkaev,
Fernando Guevara Vasquez,
China Mauck,
Milo Prisbrey,
Bart Raeymaekers
Abstract:
We theoretically show that a superposition of plane waves causes small (compared to the wavelength) particles dispersed in a fluid to assemble in quasiperiodic two or three dimensional patterns. We experimentally demonstrate this theory by using ultrasound waves to assemble quasiperiodic patterns of carbon nanoparticles in water using an octagonal arrangement of ultrasound transducers, and we docu…
▽ More
We theoretically show that a superposition of plane waves causes small (compared to the wavelength) particles dispersed in a fluid to assemble in quasiperiodic two or three dimensional patterns. We experimentally demonstrate this theory by using ultrasound waves to assemble quasiperiodic patterns of carbon nanoparticles in water using an octagonal arrangement of ultrasound transducers, and we document good agreement between theory and experiments. The theory also applies to obtaining quasiperiodic patterns in other situations where particles move with linear waves, such as optical lattices.
△ Less
Submitted 24 November, 2020;
originally announced November 2020.
-
Geometric series expansion of the Neumann-Poincaré operator: application to composite materials
Authors:
Elena Cherkaev,
Minwoo Kim,
Mikyoung Lim
Abstract:
The Neumann-Poincaré operator, a singular integral operator on the boundary of a domain, naturally appears when one solves a conductivity transmission problem via the boundary integral formulation. Recently, a series expression of the Neumann-Poincaré operator was developed in two dimensions based on geometric function theory. In this paper, we investigate geometric properties of composite materia…
▽ More
The Neumann-Poincaré operator, a singular integral operator on the boundary of a domain, naturally appears when one solves a conductivity transmission problem via the boundary integral formulation. Recently, a series expression of the Neumann-Poincaré operator was developed in two dimensions based on geometric function theory. In this paper, we investigate geometric properties of composite materials by using this series expansion. In particular, we obtain explicit formulas for the polarization tensor and the effective conductivity for an inclusion or a periodic array of inclusions of arbitrary shape with extremal conductivity, in terms of the associated exterior conformal map**. Also, we observe by numerical computations that the spectrum of the Neumann--Poincaré operator has a monotonic behavior with respect to the shape deformation of the inclusion. Additionally, we derive inequality relations of the coefficients of the Riemann map** of an arbitrary Lipschitz domain by using the properties of the polarization tensor corresponding to the domain.
△ Less
Submitted 16 September, 2020;
originally announced September 2020.
-
Homogenization of quasiperiodic structures and two-scale cut-and-projection convergence
Authors:
Niklas Wellander,
Sébastien Guenneau,
Elena Cherkaev
Abstract:
Quasiperiodic arrangements of the constitutive materials in composites result in effective properties with very unusual electromagnetic and elastic properties. The paper discusses the cut-and-projection method that is used to characterize effective properties of quasiperiodic materials. Characterization of cut-and-projection convergence limits of partial differential operators is presented, and co…
▽ More
Quasiperiodic arrangements of the constitutive materials in composites result in effective properties with very unusual electromagnetic and elastic properties. The paper discusses the cut-and-projection method that is used to characterize effective properties of quasiperiodic materials. Characterization of cut-and-projection convergence limits of partial differential operators is presented, and correctors are established. We provide the proofs of the results announced in (Wellander et al., 2018) and give further examples. Applications to problems of interest in physics include electrostatic, elastostatic and quasistatic magnetic cases.
△ Less
Submitted 8 November, 2019;
originally announced November 2019.
-
Model reduction for fractional elliptic problems using Kato's formula
Authors:
Huy Dinh,
Harbir Antil,
Yanlai Chen,
Elena Cherkaev,
Akil Narayan
Abstract:
We propose a novel numerical algorithm utilizing model reduction for computing solutions to stationary partial differential equations involving the spectral fractional Laplacian. Our approach utilizes a known characterization of the solution in terms of an integral of solutions to classical elliptic problems. We reformulate this integral into an expression whose continuous and discrete formulation…
▽ More
We propose a novel numerical algorithm utilizing model reduction for computing solutions to stationary partial differential equations involving the spectral fractional Laplacian. Our approach utilizes a known characterization of the solution in terms of an integral of solutions to classical elliptic problems. We reformulate this integral into an expression whose continuous and discrete formulations are stable; the discrete formulations are stable independent of all discretization parameters. We subsequently apply the reduced basis method to accomplish model order reduction for the integrand. Our choice of quadrature in discretization of the integral is a global Gaussian quadrature rule that we observe is more efficient than previously proposed quadrature rules. Finally, the model reduction approach enables one to compute solutions to multi-query fractional Laplace problems with order of magnitude less cost than a traditional solver.
△ Less
Submitted 19 April, 2019;
originally announced April 2019.
-
An Isoperimetric inequality for an integral operator on flat tori
Authors:
Braxton Osting,
Jeremy L. Marzuola,
Elena Cherkaev
Abstract:
We consider a class of Hilbert-Schmidt integral operators with an isotropic, stationary kernel acting on square integrable functions defined on flat tori. For any fixed kernel which is positive and decreasing, we show that among all unit-volume flat tori, the equilateral torus maximizes the operator norm and the Hilbert-Schmidt norm.
We consider a class of Hilbert-Schmidt integral operators with an isotropic, stationary kernel acting on square integrable functions defined on flat tori. For any fixed kernel which is positive and decreasing, we show that among all unit-volume flat tori, the equilateral torus maximizes the operator norm and the Hilbert-Schmidt norm.
△ Less
Submitted 9 June, 2016; v1 submitted 28 October, 2015;
originally announced October 2015.
