• 论文 • 上一篇    下一篇

波形松弛算法及其在计算流体力学中的应用

杨熙   

  1. 南京航空航天大学数学系, 南京 210016
  • 收稿日期:2012-09-04 出版日期:2013-02-15 发布日期:2013-01-22
  • 基金资助:

    国家自然科学基金(11101213)资助.

杨熙. 波形松弛算法及其在计算流体力学中的应用[J]. 计算数学, 2013, 35(1): 67-88.

Yang Xi. ON WAVEFORM RELAXATION METHODS AND ITS APPLICATION IN CFD[J]. Mathematica Numerica Sinica, 2013, 35(1): 67-88.

ON WAVEFORM RELAXATION METHODS AND ITS APPLICATION IN CFD

Yang Xi   

  1. Department of Mathematics, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China
  • Received:2012-09-04 Online:2013-02-15 Published:2013-01-22
本文介绍求解线性常系数微分代数方程组的波形松弛算法, 基于Laplace积分变换得到该算法新的收敛理论. 进一步将波形松弛算法应用于求解非定常Stokes方程, 介绍并讨论了连续时间波形松弛算法CABSOR算法和离散时间波形松弛算法DABSOR算法.
This paper provides introduction to the waveform relaxation methods for solving linear constant coefficient differential-algebraic equations (DAEs). Based on the Laplace transform, a new convergence theory for these methods is presented. Furthermore, the application of the waveform relaxation methods to the solution of time-dependent Stokes equations is studied. Specifically, continuous-time waveform relaxation methods — CABSOR and discrete-time waveform relaxation methods — DABSOR are introduced and studied.

MR(2010)主题分类: 

()
[1] Adams R A and Fournier J J F. Sobolev Spaces, The Second Edition, New York and London:Academic Press, 2003.

[2] Bai Z Z, Ng M K and Pan J Y. Alternating splitting waveform relaxation method and its successiveoverrelaxation acceleration[J]. Comput. Math. Appl., 2005, 49: 157-170.

[3] Bai Z Z, Parlett B N and Wang Z Q. On generalized successive overrelaxation methods for augmentedlinear systems[J]. Numer. Math., 2005, 102: 1-38.

[4] Bai Z Z and Yang X. Continuous-time accelerated block successive overrelaxation methods fortime-dependent Stokes equations[J]. J. Comput. Appl. Math., 2012, 236: 3265-3285.

[5] Bai Z Z and Yang X. On convergence conditions of waveform relaxation methods for lineardifferential-algebraic equations[J]. J. Comput. Appl. Math., 2011, 235: 2790-2804.

[6] Beerends R J, ter Morsche H G, van den Berg J C and van de Vrie E M. Fourier and LaplaceTransforms[M]. Cambridge: Cambridge University Press, 2003.

[7] Brenan K E, Campbell S L and Petzold L R. Numerical Solution of Initial-Value Problems inDifferential-Algebraic Equations. The Second Edition, Philadelphia: SIAM Publications, 1996.

[8] Campbell S L. One canonical form for higher index linear time-varying singular systems of differentialequations[J]. Circuits Systems Signal Process., 1983, 2: 311-326.

[9] Choudhury S R. Waveform relaxation techniques for linear and nonlinear diffusion equations[J].J. Comput. Appl. Math., 1992, 42: 253-267.

[10] Crow M L and Ilic D C. The parallel implementation of the waveform relaxation method fortransient stability simulations[J]. IEEE Trans. Power Systems, 1990, 5: 922-932.

[11] Evans L C. Partial Differential Equations[M]. Providence, Rhode Island: American MathematicalSociety, 2002.

[12] Fang W, Morari M E and Smart D. Robust VLSI circuit simulation techniques based on overlappingwaveform relaxation[J]. IEEE. Trans. Comput. Aided Design, 1995, 14: 510-518.

[13] Heywood J G and Rannacher R. Finite element approximation of the nonstationary Navier-Stokesproblem. Part IV: Error analysis for second-order time discretization[J]. SIAM J. Numer. Anal.,1990, 27: 353-384.

[14] Janssen J and Vandewalle S. On SOR waveform relaxation methods[J]. SIAM J. Numer. Anal.,1997, 34: 2456-2481.

[15] Jiang Y L. Parallel dynamic iteration methods for solving nonlinear time-periodic differentialalgebraicequations[J]. J. Comput. Math., 2003, 21: 367-374.

[16] Jiang Y L. Steady-state methods of differential-algebraic equations in circuit simulation[J]. J.Circuits Systems Comput., 2005, 14: 383-394.

[17] Jiang Y L, Chan R M M and Huang Z L. A parallel approach for computing complex eigenvalueproblems[J]. IEICE Trans. Fundam. Electron. Commun. Comput. Sci., 2000, E83-A: 2000-2008.

[18] Jiang Y L and Wing O. A note on the spectra and pseudospectra of waveform relaxation operatorsfor linear differential-algebraic equations[J]. SIAM J. Numer. Anal., 2000, 38: 186-201.

[19] Kronecker L. Algebraische Reduction der Schaaren Bilinearer Formen. Berlin: Sitzungsber. Akad.der Wiss., 1890, 763-776.

[20] Kunkel P and Mehrmann V. Canonical forms for linear differential-algebraic equations with variablecoefficients[J]. J. Comput. Appl. Math., 1994, 56: 225-259.

[21] Kunkel P and Mehrmann V. Regular solutions of nonlinear differential-algebraic equatons andtheir numerical determination[J]. Numer. Math., 1998, 79: 581-600.

[22] Kunkel P and Mehrmann V. Analysis of over- and underdetermined nonlinear differential-algebraicsystems with application to nonlinear control problems[J]. Math. Control Signal Systems, 2001,14: 233-256.

[23] Kunkel P and Mehrmann V. Characterization of classes of singular linear differential-algebraicequations[J]. Electron. J. Linear Algebra, 2005, 13: 359-386.

[24] Kunkel P and Mehrmann V and Rath W. Analysis and numerical solution of control problems indescriptor form[J]. Math. Control Signal Systems, 2001, 14: 29-61.

[25] Lelarasmee E, Ruheli A and Sangiovanni-Vincentelli A L. The waveform relaxation method fortime-domain analysis of large scale integrated circuits[J]. IEEE Trans. Comput. Aided Design,1982, CAD-1: 131-145.

[26] Lubich C. On the stability of linear multistep methods for Volterra convolution equations[J]. IMA.J. Numer. Anal., 1983, 3: 439-465.

[27] Lumsdaine A, Reichelt M W, Squyres J M and White J K. Accelerated waveform relaxationmethods for parallel transient simulation of semiconductor devices[J]. IEEE Trans. Comput. AidedDesign, 1996, 15: 716-726.

[28] Miekkala U. Dynamic iteration methods applied to linear DAE systems[J]. J. Comput. Appl.Math., 1989, 25: 133-151.

[29] Miekkala U and Nevanlinna O. Convergence of dynamic iteration methods for initial value problem[J]. SIAM J. Sci. Comput., 1987, 8: 459-482.

[30] Miekkala U and Nevanlinna O. Sets of convergence and stability regions[J]. BIT Numer. Math.,1987, 27: 554-584.

[31] Nevanlinna O. Remarks on Picard-Lindelöf iteration, Part I[J]. BIT Numer. Math., 1989, 29:328-346.

[32] Nevanlinna O. Remarks on Picard-Lindelöf iteration, Part II[J]. BIT Numer. Math., 1989, 29:535-562.

[33] Oosterlee C. Robust Multigrid Methods for the Steady and Unsteady Incompressible Navier-Stokes Equations in General Coordinates, Ph.D thesis, Department of Technical Mathematicsand Informatics, Technische Universiteit Delft, The Netherlands, 1993.

[34] Pan J Y and Bai Z Z. On the convergence of waveform relaxation methods for linear initial valueproblems[J]. J. Comput. Math., 2004, 22: 681-698.

[35] Pan J Y, Bai Z Z and Ng M K. Two-step waveform relaxation methods for implicit linear initialvalue problems[J]. Numer. Linear Algebra Appl., 2005, 12: 293-304.

[36] Rabier P J and Rheinboldt W C. Theoretical and Numerical Analysis of Differential-AlgebraicEquations, volumn VIII of Handbook of Numerical Analysis, Amsterdam: Elsevier Publicaitons,2002.

[37] Raman S, Patnaik L and Mall R. Parallel implementation of circuit simulation[J]. Intern. J. HighSpeed Comput., 1990, 2: 351-373.

[38] Rannacher R. Finite element solution of diffusion problems with irregular data[J]. Numer. Math.,1984, 43: 309-327.

[39] Reichelt M W, White J K and Allen J. Waveform relaxation for transient simulation of twodimensionalMOS devices, In ICCAD-89: IEEE International Conference on Computer-AidedDesign, Santa Clara, California, November 5-9, 1989, pages 412-415. IEEE Computer SocietyPress, Washington D.C., 1989.

[40] Rheinboldt W C. Differential-algebraic systems as differential equations on manifolds[J]. Math.Comput., 1984, 43: 473-482.

[41] Secchi A, Morari M and Biscaja E. The waveform relaxation method in the concurrent dynamicprocess simulation[J]. Comput. Chem. Engineering, 1993, 17: 683-704.

[42] Thomée V. Galerkin Finite Element Methods for Parabolic Problems. Berlin and Heidelberg:Springer-Verlag, 2006.

[43] WeierstraβK. Über ein die homogenen Funktionen zweiten Grades betreffendes Theorem, nebstAnwendung desselben auf die Theorie der kleinen Schwingungen, Berlin: Monatsh. Akad. derWissensch., 1858, 207-220.

[44] WeierstraβK. Zur Theorie der bilinearen quadratischen Fromen. Berlin: Monatsh. Akad. der Wissensch.,1867, 310-338.

[45] White J and Sangiovanni-Vincentelli A. Relaxation Techniques for the Simulation of VLSI Circuits.Boston: Kluwer Academic Publishers, 1987.

[46] Yang X. Discrete-time accelerated block successive overrelaxation methods for time-dependentStokes equations[J]. Appl. Math. Comput., submitted, 2012.
[1] 赵智慧, 李宏, 方志朝. 非定常Stokes方程的全离散稳定化有限元格式[J]. 计算数学, 2014, 36(1): 85-98.
阅读次数
全文


摘要