An economic cascadic tensor multigrid method for solving high dimensional elliptic linear partial differential problems

Chenliang Li; Jingyu Huang

arxiv: 2606.25643 · v1 · pith:GGZBIWY6 · submitted 2026-06-24 · math.NA · cs.NA

An economic cascadic tensor multigrid method for solving high dimensional elliptic linear partial differential problems

Jingyu Huang , Chenliang Li This is my paper

Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel 2026-06-25 20:45 UTCgrok-4.3pith:GGZBIWY6 record.json open to challenge →

classification math.NA cs.NA

keywords cascadic multigridtensor formhigh-dimensional elliptic PDEconjugate gradientcomputational complexitynumerical analysis

0 comments

The pith

The economic cascadic tensor multigrid method solves high-dimensional elliptic linear PDEs with O(n²) complexity instead of O(n³).

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces the ECTMG method, which applies a tensor-form representation to construct a cascadic multigrid hierarchy for high-dimensional elliptic problems. This structure lowers both storage requirements and per-iteration work so that total cost scales quadratically rather than cubically with the discretization parameter n. Convergence of the underlying tensor-based conjugate gradient solver is analyzed, and the full method is shown to inherit a comparable rate. Numerical tests confirm that the predicted scaling holds in practice. A reader would care because many scientific models produce elliptic operators in dimensions where cubic scaling quickly becomes prohibitive.

Core claim

By casting the discrete elliptic operator in tensor form, the ECTMG algorithm builds a cascadic multigrid hierarchy in which each level’s work and storage remain quadratic in the one-dimensional grid size n. The convergence rate of the tensor-form conjugate gradient iteration CG_BTF is bounded explicitly, and this bound carries over to the cascadic scheme, yielding an overall solver whose complexity is O(n²) rather than the O(n³) cost of standard dense or sparse direct methods.

What carries the argument

The cascadic tensor multigrid hierarchy (ECTMG), whose per-level operators and smoothers are kept in tensor-product form so that matrix-vector products and residual computations cost only O(n²) operations.

If this is right

Storage for the discrete operator and all iterates drops from cubic to quadratic in n.
The conjugate gradient iteration based on the tensor form (CG_BTF) converges at a rate independent of the tensor rank.
The full cascadic scheme inherits the same convergence bound and therefore terminates in a number of iterations independent of n.
Numerical examples confirm that observed runtimes follow the predicted O(n²) scaling for representative high-dimensional elliptic test problems.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

If the tensor structure survives under more general boundary conditions or variable coefficients, the same complexity reduction could apply to a wider class of linear elliptic operators.
The quadratic scaling opens the possibility of solving problems whose one-dimensional grid size n is several times larger than what cubic methods allow on the same hardware.
Because the method is cascadic rather than full multigrid, it may be combined with existing adaptive refinement strategies without changing the leading-order cost.

Load-bearing premise

The tensor representation of the discrete operator allows a cascadic hierarchy in which every level’s arithmetic and storage costs stay strictly quadratic in n without hidden cubic factors.

What would settle it

A timing experiment on successively refined grids that records total runtime growing as n³ (or worse) for large n would show that the quadratic scaling claim does not hold.

read the original abstract

In this paper, based on the tensor form, we propose a class of economic cascadic tensor multigrid method(ECTMG) for solving high dimensional elliptic linear partial differential problems. Compared with traditional methods, the new method not only reduces the storage space but also lowers the computational complexity from $\mathcal{O}(n^{3})$ to $\mathcal{O}(n^{2})$. We analyze the convergence rate of the conjugate gradient method which is based on the tensor form($\mathrm{CG}\_\mathrm{BTF}$), and then provide the convergence analysis for the new method. Finally, the effectiveness of the new method is verified through numerical examples.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The ECTMG method claims an O(n²) complexity reduction for high-dimensional elliptic PDEs via tensor cascadic multigrid, with convergence analysis, but the per-level cost control needs verification.

read the letter

The paper's main contribution is an economic cascadic tensor multigrid method, ECTMG, for high dimensional elliptic linear PDEs. It uses tensor representations to reduce both storage and computational complexity from O(n^3) to O(n^2).

The new element is the specific integration of cascadic multigrid with the tensor form of the discrete operator, along with a conjugate gradient method in tensor form called CG_BTF. The authors provide convergence analysis for CG_BTF and then for the full ECTMG method. Numerical examples are used to check the effectiveness and presumably the complexity claims.

This approach builds on existing tensor and multigrid ideas but packages them in a way that targets practical efficiency for these problems. The convergence results appear to rest on standard arguments adapted to the tensor setting, which is reasonable.

The potential weak point is the complexity reduction itself. The O(n^2) bound requires that every operation at each level, including matrix-vector products via CG_BTF, prolongations, and restrictions, remains quadratic in n without additional factors that scale with dimension or number of levels. The stress test raises a fair point here: if any tensor contraction or transfer adds even one extra n, or if CG iterations increase with n, the total cost could climb back toward cubic. The paper's analysis would need to address this explicitly to make the claim solid.

Overall, the work is for researchers in numerical PDEs who deal with high-dimensional problems and are familiar with multigrid or tensor methods. A reader looking for new algorithmic ideas in that niche could find value in the combination and the reported savings.

It deserves a serious referee because the claims are concrete and the method is testable through the numerics and analysis provided. Minor revisions might be needed to clarify the complexity details, but the core is worth reviewing.

Referee Report

2 major / 1 minor

Summary. The paper proposes an economic cascadic tensor multigrid method (ECTMG) for high-dimensional elliptic linear PDEs. Using a tensor representation of the discrete operator, it claims to reduce both storage and computational complexity from O(n³) to O(n²). The authors state that they analyze the convergence rate of the tensor-form conjugate gradient method (CG_BTF), supply convergence analysis for ECTMG, and verify effectiveness via numerical examples.

Significance. If the O(n²) complexity bound and the supporting convergence results hold with constants independent of dimension and levels, the work would address a central bottleneck in high-dimensional elliptic problems. The combination of tensor algebra with a cascadic hierarchy is a natural direction; explicit, parameter-free derivations or machine-checked bounds would strengthen the contribution.

major comments (2)

[Abstract] The central complexity claim (reduction from O(n³) to O(n²)) rests on every matvec, prolongation, and restriction in the cascadic hierarchy remaining strictly quadratic in n with no hidden linear factors from tensor contractions or inter-grid transfers. The abstract invokes this premise for both the complexity statement and the convergence analysis of CG_BTF, yet supplies none of the actual operator representations, iteration counts, or cost derivations needed to verify the bound.
[Abstract] Convergence analysis for CG_BTF is asserted but not exhibited; without the explicit error estimates, contraction factors, or dependence on the number of levels, it is impossible to confirm that the number of CG iterations remains bounded independently of n (which would be required to keep total work quadratic).

minor comments (1)

Define the acronyms CG_BTF and ECTMG at first use and state the precise tensor format of the discrete elliptic operator.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the detailed review and constructive feedback on our manuscript. We address each major comment below, clarifying where the supporting details appear in the full text and indicating revisions to improve accessibility of those details.

read point-by-point responses

Referee: [Abstract] The central complexity claim (reduction from O(n³) to O(n²)) rests on every matvec, prolongation, and restriction in the cascadic hierarchy remaining strictly quadratic in n with no hidden linear factors from tensor contractions or inter-grid transfers. The abstract invokes this premise for both the complexity statement and the convergence analysis of CG_BTF, yet supplies none of the actual operator representations, iteration counts, or cost derivations needed to verify the bound.

Authors: The abstract summarizes the main results at a high level. The tensor representations of the discrete operators, explicit forms of the matrix-vector products, prolongation and restriction operators, iteration counts for CG_BTF, and the full cost derivations establishing the O(n²) bound without hidden linear factors are derived and presented in Sections 2 (Tensor Representation of the Operator) and 4 (Complexity Analysis) of the manuscript. We will revise the abstract to include forward references to these sections. revision: partial
Referee: [Abstract] Convergence analysis for CG_BTF is asserted but not exhibited; without the explicit error estimates, contraction factors, or dependence on the number of levels, it is impossible to confirm that the number of CG iterations remains bounded independently of n (which would be required to keep total work quadratic).

Authors: The convergence analysis of CG_BTF, including the explicit error estimates, contraction factors, and proof that the iteration count remains bounded independently of n and the number of levels (under the stated assumptions on the elliptic operator), is given in Section 3 of the manuscript. This bound is used to establish the overall quadratic complexity. We will revise the abstract to point readers to Section 3 and, if needed, add a short summary sentence there for emphasis. revision: partial

Circularity Check

0 steps flagged

No circularity: derivation rests on tensor algebra properties and standard multigrid convergence arguments

full rationale

The paper defines the ECTMG method from the tensor representation of the discrete elliptic operator, then analyzes convergence of the associated CG_BTF solver and the cascadic hierarchy using standard multigrid theory and tensor contraction costs. The O(n³)→O(n²) claim follows directly from the per-level arithmetic of tensor-form matvecs, prolongations, and restrictions remaining quadratic in n, without any fitted parameter or self-referential definition that equates a prediction to its own input. No self-citation chain, uniqueness theorem imported from prior work, or ansatz smuggled via citation is invoked as load-bearing support. The numerical examples serve only as verification, not as the source of the complexity bound. The derivation is therefore self-contained against external benchmarks of tensor algebra and multigrid analysis.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Only the abstract is available; no explicit free parameters, background axioms, or newly postulated entities can be identified.

pith-pipeline@v0.9.1-grok · 5633 in / 1282 out tokens · 28514 ms · 2026-06-25T20:45:17.808233+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

32 extracted references

[1]

Solovchuk, Tony W

M A Diaz, Maxim A. Solovchuk, Tony W. H. Sheu. A conservative numerical scheme for modeling nonlinear acoustic propagation in thermoviscous homoge- neous media[J]. Journal of Computational Physics, 2018, Vol. 363: 200-230

2018
[2]

Nonlinear resonant phenomena in multilevel quantum systems[D]

H Christian. Nonlinear resonant phenomena in multilevel quantum systems[D]. Michigan State University, 2008

2008
[3]

Fast image inpainting based on coherence trans- port[J]

F Bornemann, Tom M¨ arz. Fast image inpainting based on coherence trans- port[J]. Journal of Mathematical Imaging and Vision, 2007, Vol. 28(3): 259-278

2007
[4]

Advances in the mechanics of undamageable materials: General three-dimensional formulation[J]

G Z Voyiadjis. Advances in the mechanics of undamageable materials: General three-dimensional formulation[J]. International Journal of Damage Mechanics, 2019, Vol. 28(7): 1021-1037

2019
[5]

High-order non-uniform grid schemes for numerical simulation of hypersonic boundary-level stability and transition[J]

X L Zhong , M Tatineni. High-order non-uniform grid schemes for numerical simulation of hypersonic boundary-level stability and transition[J]. Journal of Computational Physics, 2003, Vol. 190(2): 419-458

2003
[6]

JAX-Fluids: A fully-differentiable high-order computational fluid dynamics solver for compressible two-phase flows[J]

D A Bezgin, A B Buhendwa, N A Adams. JAX-Fluids: A fully-differentiable high-order computational fluid dynamics solver for compressible two-phase flows[J]. Comput. Phys. Commun., 2023, Vol. 282: 108527

2023
[7]

Wall boundary conditions for high-order finite-difference schemes in computational aeroacoustics[J]

C K W Tam, D Zhong. Wall boundary conditions for high-order finite-difference schemes in computational aeroacoustics[J]. Theoretical and Computational Fluid Dynamics, 1994, Vol. 6: 303-322

1994
[8]

Hermitian and skew-Hermitian splitting methods for non-Hermitian positive definite linear systems[J]

Z Z Bai, GH Golub, M K Ng. Hermitian and skew-Hermitian splitting methods for non-Hermitian positive definite linear systems[J]. SIAM Journal on Matrix Analysis and Applications, 2002, Vol. 24(3): 603-626

2002
[9]

Existence and Computation of Low Kronecker-Rank Approxima- tions for Large Linear Systems of Tensor Product Structure[J]

L Grasedyck. Existence and Computation of Low Kronecker-Rank Approxima- tions for Large Linear Systems of Tensor Product Structure[J]. COMPUTING, 2004, Vol. 72(3): 247-265

2004
[10]

Chebyshev collocation spectral methods for coupled radiation and conduction in a concentric spherical participating medium(Article)[J]

B W Li, Y S Sun, D W Zhang. Chebyshev collocation spectral methods for coupled radiation and conduction in a concentric spherical participating medium(Article)[J]. Journal of Heat Transfer, 2009, Vol. 131(6): 1-9. 20

2009
[11]

A projection method and Kronecker product preconditioner for solving Sylvester tensor equations[J]

Z Chen, L Z Lu. A projection method and Kronecker product preconditioner for solving Sylvester tensor equations[J]. Science China Mathematics, 2012, 55: 1281-1292

2012
[12]

On the solution to the Sylvester matrix equationAV+BW= EV F[J]

G R Duan. On the solution to the Sylvester matrix equationAV+BW= EV F[J]. IEEE Transactions on Automatic Control, 1996, 41: 612–614

1996
[13]

A complete analytical solution to the equationT A−F T=LCand its applications[J]

C C Tsui. A complete analytical solution to the equationT A−F T=LCand its applications[J]. IEEE Transactions on Automatic Control, 1987, 32: 742–744

1987
[14]

Pole assignment for linear and quadratic systems with Time-delay in control[J]

E Chu, W Lin, Reichel L. Pole assignment for linear and quadratic systems with Time-delay in control[J]. Numerical Linear Algebra with Applications, 2013, 20: 291-301

2013
[15]

Schur-decomposition for 3D matrix equations and its application in solving radiative discrete ordinates equations discretized by Chebyshev collocation spectral method[J]

B W Li, S Tian, Y S Sun. Schur-decomposition for 3D matrix equations and its application in solving radiative discrete ordinates equations discretized by Chebyshev collocation spectral method[J]. Journal of Computational Physics, 2010, 229: 1198-1212

2010
[16]

On the Krylov subspace methods based on tensor format for positive definite Sylvester tensor equations[J]

F P A Beik, F M Saberi, S Ahmadi. On the Krylov subspace methods based on tensor format for positive definite Sylvester tensor equations[J]. Numerical linear algebra with applications, 2016, 23(3): 444-466

2016
[17]

On global iterative schemes based on Hessenberg process for (ill-posed) Sylvester tensor equations[J]

M Najafi-Kalyani, F P A Beik, Jbilou K. On global iterative schemes based on Hessenberg process for (ill-posed) Sylvester tensor equations[J]. Journal of Computational and Applied Mathematics, 2020, 373: 112216

2020
[18]

A tensor format for the general- ized Hessenberg method for solving Sylvester tensor equations[J]

M Heyouni,F Saberi-Movahed, A Tajaddini. A tensor format for the general- ized Hessenberg method for solving Sylvester tensor equations[J]. Journal of Computational and Applied Mathematics, 2020, 377: 112878

2020
[19]

Researches on the iterative algorithms for several classes of matrix equations and tensor equations[D]

B H Huang. Researches on the iterative algorithms for several classes of matrix equations and tensor equations[D]. Fujian Normal University, 2019

2019
[20]

A Tensor multigrid method for solving Sylvester tensor equations[J]

Y H Chen, C L Li. A Tensor multigrid method for solving Sylvester tensor equations[J]. IEEE Transactions on Automation Science and Engineering, 2023

2023
[21]

The cascadic multigrid method for elliptic prob- lems[J]

F A Bornemann, P Deuflhard. The cascadic multigrid method for elliptic prob- lems[J]. Numerische Mathematik, 1996, 75(2): 135-152

1996
[22]

A new class of cascadic multigrid method, Sci

Z C Shi, X J Xu. A new class of cascadic multigrid method, Sci. China Math. 30(09)(2000), 799–807. 21

2000
[23]

Z C Shi, X J Xu, Y. Q. Huang. Economical cascadic multigrid method (ECMG), Sci. China Math. 37(09) (2007), 1083–1098

2007
[24]

A modulus-based cascadic multigrid method for elliptic vari- ational inequality problems[J]

Y Wang, C L Li. A modulus-based cascadic multigrid method for elliptic vari- ational inequality problems[J]. Numerical Algorithms, 2022, 90(4): 1777–1791

2022
[25]

Modulus-based cascadic multigrid method for quasi-variational inequality problems[J]

K Y Gao, C L Li. Modulus-based cascadic multigrid method for quasi-variational inequality problems[J]. Communications on Applied Mathematics and Compu- tation, 2024, 6(4)

2024
[26]

Developing iterative algorithms to solve Sylvester tensor equations[J]

X F Zhang, Q W Wang. Developing iterative algorithms to solve Sylvester tensor equations[J]. Applied Mathematics and Computation, 2021, 409: 126403

2021
[27]

Kolda, Brett W

Tamara G. Kolda, Brett W. Bader. Tensor Decompositions and Applications[J]. SIAM Review, 2009, Vol. 51(3): 455-500

2009
[28]

A projection method to solve linear system in tensor format[J]

J Ballani, L Grasedyck. A projection method to solve linear system in tensor format[J]. Numerical Linear Algebra with Applications, 2013, 20(1): 27-43

2013
[29]

Solution of linear systems in high spatial dimensions[J]

W Hackbusch. Solution of linear systems in high spatial dimensions[J]. Com- puting and visualization in science, 2015, 17(3): 111-118

2015
[30]

Mathematical foundation of finite element method[M]

L H Wang, X J Xu. Mathematical foundation of finite element method[M]. Beijing: Science Press, 2004: 54-64

2004
[31]

Finite difference methods for ordinary and partial differential equations: steady-state and time-dependent problems[M]

R J LeVeque. Finite difference methods for ordinary and partial differential equations: steady-state and time-dependent problems[M]. Godin Lyttle press, 2009

2009
[32]

Research on the meshless solving algorithm for 3D steady convection-diffusion problems[J]

X H Zhang, J H Deng. Research on the meshless solving algorithm for 3D steady convection-diffusion problems[J]. Applied Mathematics and Mechanics, 2014, 35(11): 1249-1258. 22

2014

[1] [1]

Solovchuk, Tony W

M A Diaz, Maxim A. Solovchuk, Tony W. H. Sheu. A conservative numerical scheme for modeling nonlinear acoustic propagation in thermoviscous homoge- neous media[J]. Journal of Computational Physics, 2018, Vol. 363: 200-230

2018

[2] [2]

Nonlinear resonant phenomena in multilevel quantum systems[D]

H Christian. Nonlinear resonant phenomena in multilevel quantum systems[D]. Michigan State University, 2008

2008

[3] [3]

Fast image inpainting based on coherence trans- port[J]

F Bornemann, Tom M¨ arz. Fast image inpainting based on coherence trans- port[J]. Journal of Mathematical Imaging and Vision, 2007, Vol. 28(3): 259-278

2007

[4] [4]

Advances in the mechanics of undamageable materials: General three-dimensional formulation[J]

G Z Voyiadjis. Advances in the mechanics of undamageable materials: General three-dimensional formulation[J]. International Journal of Damage Mechanics, 2019, Vol. 28(7): 1021-1037

2019

[5] [5]

High-order non-uniform grid schemes for numerical simulation of hypersonic boundary-level stability and transition[J]

X L Zhong , M Tatineni. High-order non-uniform grid schemes for numerical simulation of hypersonic boundary-level stability and transition[J]. Journal of Computational Physics, 2003, Vol. 190(2): 419-458

2003

[6] [6]

JAX-Fluids: A fully-differentiable high-order computational fluid dynamics solver for compressible two-phase flows[J]

D A Bezgin, A B Buhendwa, N A Adams. JAX-Fluids: A fully-differentiable high-order computational fluid dynamics solver for compressible two-phase flows[J]. Comput. Phys. Commun., 2023, Vol. 282: 108527

2023

[7] [7]

Wall boundary conditions for high-order finite-difference schemes in computational aeroacoustics[J]

C K W Tam, D Zhong. Wall boundary conditions for high-order finite-difference schemes in computational aeroacoustics[J]. Theoretical and Computational Fluid Dynamics, 1994, Vol. 6: 303-322

1994

[8] [8]

Hermitian and skew-Hermitian splitting methods for non-Hermitian positive definite linear systems[J]

Z Z Bai, GH Golub, M K Ng. Hermitian and skew-Hermitian splitting methods for non-Hermitian positive definite linear systems[J]. SIAM Journal on Matrix Analysis and Applications, 2002, Vol. 24(3): 603-626

2002

[9] [9]

Existence and Computation of Low Kronecker-Rank Approxima- tions for Large Linear Systems of Tensor Product Structure[J]

L Grasedyck. Existence and Computation of Low Kronecker-Rank Approxima- tions for Large Linear Systems of Tensor Product Structure[J]. COMPUTING, 2004, Vol. 72(3): 247-265

2004

[10] [10]

Chebyshev collocation spectral methods for coupled radiation and conduction in a concentric spherical participating medium(Article)[J]

B W Li, Y S Sun, D W Zhang. Chebyshev collocation spectral methods for coupled radiation and conduction in a concentric spherical participating medium(Article)[J]. Journal of Heat Transfer, 2009, Vol. 131(6): 1-9. 20

2009

[11] [11]

A projection method and Kronecker product preconditioner for solving Sylvester tensor equations[J]

Z Chen, L Z Lu. A projection method and Kronecker product preconditioner for solving Sylvester tensor equations[J]. Science China Mathematics, 2012, 55: 1281-1292

2012

[12] [12]

On the solution to the Sylvester matrix equationAV+BW= EV F[J]

G R Duan. On the solution to the Sylvester matrix equationAV+BW= EV F[J]. IEEE Transactions on Automatic Control, 1996, 41: 612–614

1996

[13] [13]

A complete analytical solution to the equationT A−F T=LCand its applications[J]

C C Tsui. A complete analytical solution to the equationT A−F T=LCand its applications[J]. IEEE Transactions on Automatic Control, 1987, 32: 742–744

1987

[14] [14]

Pole assignment for linear and quadratic systems with Time-delay in control[J]

E Chu, W Lin, Reichel L. Pole assignment for linear and quadratic systems with Time-delay in control[J]. Numerical Linear Algebra with Applications, 2013, 20: 291-301

2013

[15] [15]

Schur-decomposition for 3D matrix equations and its application in solving radiative discrete ordinates equations discretized by Chebyshev collocation spectral method[J]

B W Li, S Tian, Y S Sun. Schur-decomposition for 3D matrix equations and its application in solving radiative discrete ordinates equations discretized by Chebyshev collocation spectral method[J]. Journal of Computational Physics, 2010, 229: 1198-1212

2010

[16] [16]

On the Krylov subspace methods based on tensor format for positive definite Sylvester tensor equations[J]

F P A Beik, F M Saberi, S Ahmadi. On the Krylov subspace methods based on tensor format for positive definite Sylvester tensor equations[J]. Numerical linear algebra with applications, 2016, 23(3): 444-466

2016

[17] [17]

On global iterative schemes based on Hessenberg process for (ill-posed) Sylvester tensor equations[J]

M Najafi-Kalyani, F P A Beik, Jbilou K. On global iterative schemes based on Hessenberg process for (ill-posed) Sylvester tensor equations[J]. Journal of Computational and Applied Mathematics, 2020, 373: 112216

2020

[18] [18]

A tensor format for the general- ized Hessenberg method for solving Sylvester tensor equations[J]

M Heyouni,F Saberi-Movahed, A Tajaddini. A tensor format for the general- ized Hessenberg method for solving Sylvester tensor equations[J]. Journal of Computational and Applied Mathematics, 2020, 377: 112878

2020

[19] [19]

Researches on the iterative algorithms for several classes of matrix equations and tensor equations[D]

B H Huang. Researches on the iterative algorithms for several classes of matrix equations and tensor equations[D]. Fujian Normal University, 2019

2019

[20] [20]

A Tensor multigrid method for solving Sylvester tensor equations[J]

Y H Chen, C L Li. A Tensor multigrid method for solving Sylvester tensor equations[J]. IEEE Transactions on Automation Science and Engineering, 2023

2023

[21] [21]

The cascadic multigrid method for elliptic prob- lems[J]

F A Bornemann, P Deuflhard. The cascadic multigrid method for elliptic prob- lems[J]. Numerische Mathematik, 1996, 75(2): 135-152

1996

[22] [22]

A new class of cascadic multigrid method, Sci

Z C Shi, X J Xu. A new class of cascadic multigrid method, Sci. China Math. 30(09)(2000), 799–807. 21

2000

[23] [23]

Z C Shi, X J Xu, Y. Q. Huang. Economical cascadic multigrid method (ECMG), Sci. China Math. 37(09) (2007), 1083–1098

2007

[24] [24]

A modulus-based cascadic multigrid method for elliptic vari- ational inequality problems[J]

Y Wang, C L Li. A modulus-based cascadic multigrid method for elliptic vari- ational inequality problems[J]. Numerical Algorithms, 2022, 90(4): 1777–1791

2022

[25] [25]

Modulus-based cascadic multigrid method for quasi-variational inequality problems[J]

K Y Gao, C L Li. Modulus-based cascadic multigrid method for quasi-variational inequality problems[J]. Communications on Applied Mathematics and Compu- tation, 2024, 6(4)

2024

[26] [26]

Developing iterative algorithms to solve Sylvester tensor equations[J]

X F Zhang, Q W Wang. Developing iterative algorithms to solve Sylvester tensor equations[J]. Applied Mathematics and Computation, 2021, 409: 126403

2021

[27] [27]

Kolda, Brett W

Tamara G. Kolda, Brett W. Bader. Tensor Decompositions and Applications[J]. SIAM Review, 2009, Vol. 51(3): 455-500

2009

[28] [28]

A projection method to solve linear system in tensor format[J]

J Ballani, L Grasedyck. A projection method to solve linear system in tensor format[J]. Numerical Linear Algebra with Applications, 2013, 20(1): 27-43

2013

[29] [29]

Solution of linear systems in high spatial dimensions[J]

W Hackbusch. Solution of linear systems in high spatial dimensions[J]. Com- puting and visualization in science, 2015, 17(3): 111-118

2015

[30] [30]

Mathematical foundation of finite element method[M]

L H Wang, X J Xu. Mathematical foundation of finite element method[M]. Beijing: Science Press, 2004: 54-64

2004

[31] [31]

Finite difference methods for ordinary and partial differential equations: steady-state and time-dependent problems[M]

R J LeVeque. Finite difference methods for ordinary and partial differential equations: steady-state and time-dependent problems[M]. Godin Lyttle press, 2009

2009

[32] [32]

Research on the meshless solving algorithm for 3D steady convection-diffusion problems[J]

X H Zhang, J H Deng. Research on the meshless solving algorithm for 3D steady convection-diffusion problems[J]. Applied Mathematics and Mechanics, 2014, 35(11): 1249-1258. 22

2014