SearcharxivSearch

arXiv subjects

Yangwen Zhang

Publications and source records attributed to Yangwen Zhang.

At least 19 recordsLinked to original sources

Sharp and unified $L^2$ error estimates for the nonsymmetric Nitsche method on convex polytopes

Nitsche's method weakly imposes Dirichlet boundary conditions, but its nonsymmetric variant has long shown a gap between theory and computation: the classical $L^2$~analysis under $H^{k+1}$~regularity predicts a half-order convergence loss, whereas numerical experiments on smooth test problems consistently produce the optimal rate. Whether this discrepancy reflects a limitation of the analysis or an essential feature of the method has remained an open question. On bounded convex polytopes in two and three dimensions, we prove a unified, regularity-dependent $L^2$~error estimate valid across the entire penalty scale $h^{-α}$: \begin{align*} \|u-u_h\|_{L^2(Ω)}\le C h^r |u|_{W^{k+1,p}(Ω)},\qquad r=\min\bigl\{k+1,\,k+\max\{1,α\}-1/p\bigr\}. \end{align*} Numerical experiments in two and three dimensions, on a one-parameter family of manufactured solutions with tunable regularity, demonstrate the sharpness of the estimate and resolve the open question. First, under merely $H^{k+1}$~regularity the half-order loss is essential; second, the optimal convergence consistently observed on smooth test problems is therefore explained by their full $W^{k+1,\infty}$~regularity, not by a limitation of the standard analysis. The theory identifies $α\ge 1+1/p$ or $p=\infty$ as the sharp threshold for recovering the optimal rate $h^{k+1}$, and the experiments confirm this if-and-only-if condition in both dimensions.

math.NA

A fast reduced order method for linear parabolic inverse source problems

In this paper, we propose a novel, computationally efficient reduced order method to solve linear parabolic inverse source problems. Our approach provides accurate numerical solutions without relying on specific training data. The forward solution is constructed using a Krylov sequence, while the source term is recovered via the conjugate gradient (CG) method. Under a weak regularity assumption on the solution of the parabolic partial differential equations (PDEs), we establish convergence of the forward solution and provide a rigorous error estimate for our method. Numerical results demonstrate that our approach offers substantial computational savings compared to the traditional finite element method (FEM) and retains equivalent accuracy.

math.NA

An Improved Incremental Singular Value Decomposition and New Error Bounds

The incremental singular value decomposition (SVD) updates a truncated SVD as new columns arrive, replacing a single large SVD with a sequence of small ones. In floating-point arithmetic, each update multiplies the running singular basis by a small orthogonal factor, and the accumulated product loses orthogonality unless the basis is reorthogonalized periodically. How often this reorthogonalization is needed has been an open question; we answer it by restructuring the algorithm so that rank-preserving updates are accumulated implicitly and applied in batches, reducing the number of large orthogonal multiplications from $n$, the stream length, to $r$, the numerical rank. We prove that this restructuring preserves the exact-arithmetic output of the original algorithm and establish two forward-error bounds. First, we sharpen the existing operator-norm truncation bound from $n\,\texttt{tol}$ to $\sqrt{n}\,\texttt{tol}$, and show the new rate is attained on a constructive example. Second, under a standard probabilistic rounding-error model, we prove that the loss of orthogonality of the computed left factor is independent of the stream length $n$ and depends on $m$, the length of each incoming column, only through a single $\sqrt{m}$ factor. Numerical experiments confirm both bounds and demonstrate that the proposed algorithm runs $4.5\times$ to $34\times$ faster than its closest competitors.

math.NA

A superconvergent hybridizable discontinuous Galerkin method for the convective Cahn--Hilliard equation

We propose a hybridizable discontinuous Galerkin (HDG) method combined with convex-concave splitting for the temporal discretization of the convective Cahn-Hilliard equation. The convection term is discretized explicitly without stabilization, yielding three key advantages: (1) unconditional stability, (2) preservation of the optimal convergence rate for piecewise constant approximations, and (3) a symmetric system after local elimination, enabling efficient solver via minimal residual methods. We establish optimal convergence rates in the $L^2$ norm for both the scalar and flux variables for any polynomial degree $k \geq 0$. To achieve optimal $L^2$-norm estimates, we introduce a specialized HDG elliptic projection operator and analyze its approximation properties. Within the HDG framework, local elimination is employed to reduce the degrees of freedom associated with the globally coupled unknowns, and the scalar variables exhibit superconvergence. Finally, numerical experiments validate the theoretical convergence rates and demonstrate the effectiveness of the proposed method.

math.NA

An efficient and memory free algorithm for subdiffusion equation using incremental singular value decomposition

In this paper, we address the well-known challenge in the numerical solution of time-fractional partial differential equations (TFPDEs), namely, that the dependence on all previous time levels leads to storage requirements that grow linearly with the number of time steps. To overcome this difficulty, we develop an efficient algorithm based on incremental singular value decomposition (ISVD), which avoids the excessive memory demands associated with storing the full solution history. A rigorous error analysis is established, and numerical experiments are presented to validate the theoretical results. Comparisons with the direct method and a representative fast evaluation method show that the proposed ISVD approach dramatically reduces memory usage relative to the direct method and remains competitive with the fast method over the tested parameter regimes.

math.NA

Incremental SVD Compression for Nonlinear Oldroyd Equations with General Memory Kernels

We study mixed finite element/Crank--Nicolson discretizations of a nonlinear Oldroyd problem with general nonsingular and weakly singular memory kernels. Direct evaluation of the history term requires storing all previous velocity snapshots, which leads to $\mathcal{O}(mN)$ memory and $\mathcal{O}(mN^2)$ work over $N$ time steps, where $m$ denotes the number of spatial degrees of freedom. To reduce this burden, we compress the velocity history online by an incremental singular value decomposition and use the compressed representation in the discrete memory term. Under an approximate low-rank assumption of numerical rank $r$, the storage decreases to $\mathcal{O}((m+N)r)$, while the total history-evaluation work becomes $\mathcal{O}(mNr+rN^2)$. For nonsingular kernels, we derive a tolerance-dependent perturbation estimate showing that the baseline finite element accuracy is retained when the compression tolerance is sufficiently small. We also extend the approach to tempered weakly singular kernels via convolution quadrature. Numerical tests show near-indistinguishable solutions from the uncompressed scheme for the reported tolerances, together with substantial memory savings and reduced wall-clock time.

math.NA

Pressure-Robust $H(\mathrm{div})$-Conforming HDG Methods for the Steady Stokes Equations with an Application to Tangential Boundary Control

We develop a family of $H(\mathrm{div})$-conforming hybridizable discontinuous Galerkin methods for the steady Stokes equations based on BDM and RT velocity spaces with either discontinuous or continuous hybrid traces. In contrast to our earlier pressure-robust HDG method for tangential boundary control, the present analysis does not require the pressure to belong to $H^1$; instead, the consistency argument only assumes low pressure regularity. The discrete velocities are exactly divergence-free, which yields pressure robustness. For the BDM variants we derive optimal energy-norm estimates and optimal $L^2$-velocity convergence, while for the RT variants we obtain optimal velocity convergence and weaker pressure estimates. We also analyze the hybridized linear system and prove a uniform spectral equivalence for the pressure Schur complement relevant to iterative solvers. As an application, we revisit the Stokes tangential boundary control problem and derive error estimates for the control, state, and adjoint variables using the BDM discontinuous-trace scheme. Two- and three-dimensional numerical experiments confirm the predicted convergence rates, the exact divergence-free property, and the robustness of the method with respect to the viscosity parameter.

math.NA

Analysis of a hybridizable discontinuous Galerkin scheme for the tangential control of the Stokes system

We consider an unconstrained tangential Dirichlet boundary control problem for the Stokes equations with an $ L^2 $ penalty on the boundary control. The contribution of this paper is twofold. First, we obtain well-posedness and regularity results for the tangential Dirichlet control problem on a convex polygonal domain. The analysis contains new features not found in similar Dirichlet control problems for the Poisson equation; an interesting result is that the optimal control has higher local regularity on the individual edges of the domain compared to the global regularity on the entire boundary. Second, we propose and analyze a hybridizable discontinuous Galerkin (HDG) method to approximate the solution. For convex polygonal domains, our theoretical convergence rate for the control is optimal with respect to the global regularity on the entire boundary. We present numerical experiments to demonstrate the performance of the HDG method.

math.NA

An Incremental SVD Method for Non-Fickian Flows in Porous Media: Addressing Storage and Computational Challenges

It is well known that the numerical solution of the Non-Fickian flows at the current stage depends on all previous time instances. Consequently, the storage requirement increases linearly, while the computational complexity grows quadratically with the number of time steps. This presents a significant challenge for numerical simulations. While numerous existing methods address this issue, our proposed approach stems from a data science perspective and maintains uniformity. Our method relies solely on the rank of the solution data, dissociating itself from dependency on any specific partial differential equation (PDE). In this paper, we make the assumption that the solution data exhibits approximate low rank. Here, we present a memory-free algorithm, based on the incremental SVD technique, that exhibits only linear growth in computational complexity as the number of time steps increases. We prove that the error between the solutions generated by the conventional algorithm and our innovative approach lies within the scope of machine error. Numerical experiments are showcased to affirm the accuracy and efficiency gains in terms of both memory usage and computational expenses.

math.NA

A New Global Divergence Free and Pressure-Robust HDG Method for Tangential Boundary Control of Stokes Equations

In [ESAIM: M2AN, 54(2020), 2229-2264], we proposed an HDG method to approximate the solution of a tangential boundary control problem for the Stokes equations and obtained an optimal convergence rate for the optimal control {that reflects its global regularity}. However, the error estimates depend on the pressure, and the velocity is not divergence free. The importance of pressure-robust numerical methods for fluids was addressed by John et al. [SIAM Review, 59(2017), 492-544]. In this work, we devise a new HDG method to approximate the solution of the Stokes tangential boundary control problem; the HDG method is also of independent interest for solving the Stokes equations. This scheme yields a $\mathbf{H}(\mathrm{div})$ conforming, globally divergence free, and pressure-robust solution. To the best of our knowledge, this is the first time such a numerical scheme has been obtained for an optimal boundary control problem for the Stokes equations. We also provide numerical experiments to show the performance of the new HDG method and the advantage over the non pressure-robust scheme.

math.NA

A new reduced order model of linear parabolic PDEs

How to build an accurate reduced order model (ROM) for multidimensional time dependent partial differential equations (PDEs) is quite open. In this paper, we propose a new ROM for linear parabolic PDEs. We prove that our new method can be orders of magnitude faster than standard solvers, and is also much less memory intensive. Under some assumptions on the problem data, we prove that the convergence rates of the new method is the same with standard solvers. Numerical experiments are presented to confirm our theoretical result.

math.NA

A new reduced order model of imcompressible Stokes equations

In this paper we propose a new reduced order model (ROM) to the imcompressible Stokes equations. Numerical experiments show that our ROM is accurate and efficient. Under some assumptions on the problem data, we prove that the convergence rates of the new ROM is the same with standard solvers.

math.NA

A Spectral Target Signature for Thin Surfaces with Higher Order Jump Conditions

In this paper we consider the inverse problem of determining structural properties of a thin anisotropic and dissipative inhomogeneity in ${\mathbb R}^m$, $m=2,3$ from scattering data. In the asymptotic limit as the thickness goes to zero, the thin inhomogeneity is modeled by an open $m-1$ dimensional manifold (here referred to as screen), and the field inside is replaced by jump conditions on the total field involving a second order surface differential operator. We show that all the surface coefficients (possibly matrix valued and complex) are uniquely determined from far field patterns of the scattered fields due to infinitely many incident plane waves at a fixed frequency. Then we introduce a target signature characterized by a novel eigenvalue problem such that the eigenvalues can be determined from measured scattering data, adapting the approach in \cite{Screens}. Changes in the measured eigenvalues are used to identified changes in the coefficients without making use of the governing equations that model the healthy screen. In our investigation the shape of the screen is known, since it represents the object being evaluated. We present some preliminary numerical results indicating the validity of our inversion approach.

math.AP

Sharp $L^\infty$ estimates of HDG methods for Poisson equation II: 3D

In [SIAM J. Numer. Anal., 59 (2), 720-745], we proved quasi-optimal $L^\infty$ estimates (up to logarithmic factors) for the solution of Poisson's equation by a hybridizable discontinuous Galerkin (HDG) method. However, the estimates only work in 2D. In this paper, we obtain sharp (without logarithmic factors) $L^\infty$ estimates for the HDG method in both 2D and 3D. Numerical experiments are presented to confirm our theoretical result.

math.NA

Target signatures for thin surfaces

We investigate an inverse scattering problem for a thin inhomogeneous scatterer in R^m, m = 2,3, which we model as a m-1 dimensional open surface. The scatterer is referred to as a screen. The goal is to design target signatures that are computable from scattering data in order to detect changes in the material properties of the screen. This target signature is characterized by a mixed Steklov eigenvalue problem for a domain whose boundary contains the screen. We show that the corresponding eigenvalues can be determined from appropriately modified scattering data by using the generalized linear sampling method. A weaker justification is provided for the classical linear sampling method. Numerical experiments are presented to support our theoretical results.

math.NA

Superconvergent Interpolatory HDG methods for reaction diffusion equations II: HHO-inspired methods

In J. Sci. Comput., 81: 2188-2212, 2019, we considered a superconvergent hybridizable discontinuous Galerkin (HDG) method, defined on simplicial meshes, for scalar reaction diffusion equations and showed how to define an interpolatory version which maintained its convergence properties. The interpolatory approach uses a locally postprocessed approximate solution to evaluate the nonlinear term, and assembles all HDG matrices once before the time intergration leading to a reduction in computational cost. The resulting method displays a superconvergent rate for the solution for polynomial degree $k\ge 1$. In this work, we take advantage of the link found between the HDG and the hybrid high-order (HHO) methods, in ESAIM Math. Model. Numer. Anal., 50: 635-650, 2016 and extend this idea to the new, HHO-inspired HDG methods, defined on meshes made of general polyhedral elements, uncovered therein}. We prove that the resulting interpolatory HDG methods converge at the same rate as for the linear elliptic problems. Hence, we obtain superconvergent methods for $k\ge 0$ by some methods. We present numerical results to illustrate the convergence theory.

math.NA

$L^\infty$ norm error estimates for HDG methods applied to the Poisson equation with an application to the Dirichlet boundary control problem

We prove quasi-optimal $L^\infty$ norm error estimates (up to logarithmic factors) for the solution of Poisson's problem by the standard Hybridizable Discontinuous Galerkin (HDG) method. Although such estimates are available for conforming and mixed finite element methods, this is the first proof for HDG. The method of proof is motivated by known $L^\infty$ norm estimates for mixed finite elements. We show two applications: the first is to prove optimal convergence rates for boundary flux estimates, and the second is to prove that numerically observed convergence rates for the solution of a Dirichlet boundary control problem are to be expected theoretically. Numerical examples show that the predicted rates are seen in practice.

math.NA