一种针对大波数Helmholtz方程的高性能并行预条件迭代求解算法

针对传统串行迭代法求解大波数Helmholtz方程存在效率低下且受限于单机内存的问题,提出了一种基于消息传递接口(Message Passing Interface,MPI) 的并行预条件迭代法。该算法利用复移位拉普拉斯算子对Helmholtz方程进行预条件处理,联合稳定双共轭梯度法和基于矩阵的多重网格法来求解预条件方程离散后的大规模线性系统,在Linux集群系统上基于 MPI环境实现了求解算法的并行计算,重点解决了多重网格的并行划分、信息传递和多重网格组件的构建问题。数值实验表明,对于大波数问题,提出的算法具有良好的并行加速比,相较于串行算法极大地提高了计算效率。     

Modelling variable density flow problems in heterogeneous porous media using the method of lines and advanced spatial discretization methods

Modelling variable density flow problems under heterogeneous porous media conditions requires very long computation time and high performance equipments. In this work, the DASPK solver for temporal resolution is combined with advanced spatial discretization schemes in order to improve the computational efficiency while maintaining accuracy.The spatial discretization is based on a combination of Mixed Finite Element (MFE), Discontinuous Galerkin (DG) and Multi-point Flux Approximation methods (MPFA). The obtained non-linear ODE/DAE system is solved with the Method of Lines (MOL) using the DASPK time solver. DASPK uses the preconditioned Krylov iterative method to solve linear systems arising at each time step.Precise laboratory-scale 2D experiments were conducted in a heterogeneously packed porous medium flow tank and the measured concentration contour lines are used to evaluate the numerical model. Simulations show the high efficiency and accuracy of the code and the sensitivity analysis confirms the density dependence of dispersion.     

High Performance Computations for Large Scale Simulations of Subsurface Multiphase Fluid and Heat Flow

TOUGH2 is a widely used reservoir simulator for solving subsurface flow related problems such as nuclear waste geologic isolation, environmental remediation of soil and groundwater contamination, and geothermal reservoir engineering. It solves a set of coupled mass and energy balance equations using a finite volume method. This contribution presents the design and analysis of a parallel version of TOUGH2. The parallel implementation first partitions the unstructured computational domain. For each time step, a set of coupled non-linear equations is solved with Newton iteration. In each Newton step, a Jacobian matrix is calculated and an ill-conditioned non-symmetric linear system is solved using a preconditioned iterative solver. Communication is required for convergence tests and data exchange across partitioning borders. Parallel performance results on Cray T3E-900 are presented for two real application problems arising in the Yucca Mountain nuclear waste site study. The execution time is reduced from 7504 seconds on two processors to 126 seconds on 128 processors for a 2D problem involving 52,752 equations. For a larger 3D problem with 293,928 equations the time decreases from 10,055 seconds on 16 processors to 329 seconds on 512 processors.     

Comparing direct and iterative equation solvers in a large structural analysis software system

This paper describes two vectorized implementations of preconditioned conjugate gradient (PCG) solvers. Sparse and diagonal matrix storage schemes are described and compared. A vectorized incomplete Choleski preconditioning is described and compared with Jacobi preconditioning. A modification to the basic no-fill incomplete Choleski method to improve performance and robustness is described. The two PCG solvers are compared with direct Choleski methods using a sparse Choleski solver from SPARSPAK and a vectorized variable-band Choleski solver developed at NASA Langley Research Center. All of the linear equation solvers are implemented in a large structural analysis finite element software system called the Computational Structural Mechanics (CSM) Testbed. The CSM Testbed is used to provide a common software system in which new methods are developed and tested. Several representative two- and three-dimensional structural analysis problems are solved using the various equation solvers. Results are given from runs made on the CONVEX C220 and CRAY 2 computer systems. Comparisons of the convergence rates for the iterative solvers as well as the computation rates, number of operations, and overall CPU time required by all of the equation solvers are given.     

Preconditioned variational methods on rotated finite difference discretisation

Due to their rapid convergence properties, recent focus on iterative methods in the solution of linear system has seen a flourish on the use of gradient techniques which are primarily based on global minimisation of the residual vectors. In this paper, we conduct an experimental study to investigate the performance of several preconditioned gradient or variational techniques to solve a system arising from the so-called rotated (skewed) finite difference discretisation in the solution of elliptic partial differential equations (PDEs). The preconditioned iterative methods consist of variational accelerators, namely the steepest descent and conjugate gradient methods, applied to a special matrix ‘splitting’ preconditioned system. Several numerical results are presented and discussed.     

Parallel Semi-Implicit Spectral Element Methods for Atmospheric General Circulation Models

Spectral elements combine the accuracy and exponential convergence of conventional spectral methods with the geometric flexibility of finite elements. Additionally, there are several apparent computational advantages to using spectral element methods on microprocessors. In particular, the computations are naturally cache-blocked and derivatives may be computed using nearest neighbor communications. Thus, an explicit spectral element atmospheric model has demonstrated close to linear scaling on a variety of distributed memory computers including the IBM SP and Linux Clusters. Explicit formulations of PDE's arising in geophysical fluid dynamics, such as the primitive equations on the sphere, are time-step limited by the phase speed of gravity waves. Semi-implicit time integration schemes remove the stability restriction but require the solution of an elliptic BVP. By employing a weak formulation of the governing equations, it is possible to obtain a symmetric Helmholtz operator that permits the solution of the implicit problem using conjugate gradients. We find that a block-Jacobi preconditioned conjugate gradient solver accelerates the simulation rate of the semi-implicit relative to the explicit formulation for practical climate resolutions by about a factor of three.     

Application of an element-by-element BiCGSTAB iterative solver to a monotonic finite element model

This paper deals with the advection-diffusion equation in adaptive meshes. The main feature of the present finite element model is the use of Legendre-polynomials to span finite element spaces. The success that this model gives good resolutions to solutions in regions of boundary and interior layers lies in the use of M-matrix theory. In the monotonic range of Peclet numbers, the Petrov-Galerkin method performs well in the sense that oscillatory solutions are not present in the flow. With proper stabilization, finite element matrix equations can be iteratively solved by the Lanczos method, used concurrently with local minimization provided by GMRES(1). The resulting BiCGSTAB iterative solver, supplemented with the Jacobi preconditioner, is implemented in an element-by-element fashion. This gives solutions which are computationally feasible for large-scale flow simulations. The results of two computations are presented in support of the ability of the present finite element model to resolve sharp gradients in the solution. As is apparent from this study is that considerable savings in computer storage and execution time are achieved in adaptive meshes through use of the preconditioned BiCGSTAB iterative solver.     

A Fast Preconditioned Iterative Algorithm for the Electromagnetic Scattering from a Large Cavity

In this paper, a fast preconditioned Krylov subspace iterative algorithm is proposed for the electromagnetic scattering from a rectangular large open cavity embedded in an infinite ground plane. The scattering problem is described by the Helmholtz equation with a nonlocal artificial boundary condition on the aperture of the cavity and Dirichlet boundary conditions on the walls of the cavity. Compact fourth order finite difference schemes are employed to discretize the bounded domain problem. A much smaller interface discrete system is reduced by introducing the discrete Fourier transformation in the horizontal and a Gaussian elimination in the vertical direction, presented in Bao and Sun (SIAM J. Sci. Comput. 27:553, 2005). An effective preconditioner is developed for the Krylov subspace iterative solver to solve this interface system. Numerical results demonstrate the remarkable efficiency and accuracy of the proposed method.     

An inexact interior-point method for system analysis

In this article, a primal-dual interior-point algorithm for semidefinite programming that can be used for analysing e.g. polytopic linear differential inclusions is tailored in order to be more computationally efficient. The key to the speedup is to allow for inexact search directions in the interior-point algorithm. These are obtained by aborting an iterative solver for computing the search directions prior to convergence. A convergence proof for the algorithm is given. Two different preconditioners for the iterative solver are proposed. The speedup is in many cases more than an order of magnitude. Moreover, the proposed algorithm can be used to analyse much larger problems as compared to what is possible with off-the-shelf interior-point solvers.     

On semi-convergence of parameterized SHSS method for a class of singular complex symmetric linear systems

In this paper, we use the parameterized single-step HSS (P-SHSS) iterative method to solve a broad class of singular complex symmetric linear systems. The semi-convergence properties of the P-SHSS method are derived under suitable conditions. Moreover, some properties of the preconditioned matrix and the optimal parameters are analyzed in detail. Numerical experiments are given to support our theoretical results and show the effectiveness of the P-SHSS method either as a solver or as a preconditioner.     

On the Solution of Boundary Value Problems by Using Fast Generalized Approximate Inverse Banded Matrix Techniques

A class of finite difference schemes in conjunction with approximate inverse banded matrix techniques based on the concept of LU-type factorization procedures is introduced for computing fast explicit approximate inverses. Explicit preconditioned iterative schemes in conjunction with approximate inverse matrix techniques are presented for the efficient solution of banded linear systems. A theorem on the rate of convergence and estimates of the computational complexity required to reduce the L-norm of the error is presented. Applications of the method on linear and non-linear systems are discussed and numerical results are given.     

Explicit preconditioned domain decomposition schemes for solving nonlinear boundary value problems

A new class of inner-outer iterative procedures in conjunction with Picard-Newton methods based on explicit preconditioning iterative methods for solving nonlinear systems is presented. Explicit preconditioned iterative schemes, based on the explicit computation of a class of domain decomposition generalized approximate inverse matrix techniques are presented for the efficient solution of nonlinear boundary value problems on multiprocessor systems. Applications of the new composite scheme on characteristic nonlinear boundary value problems are discussed and numerical results are given.     

A general iterative sparse linear solver and its parallelization for interval Newton methods

Interval Newton/Generalized Bisection methods reliably find all numerical solutions within a given domain. Both computational complexity analysis and numerical experiments have shown that solving the corresponding interval linear system generated by interval Newton's methods can be computationally expensive (especially when the nonlinear system is large). In applications, many large-scale nonlinear systems of equations result in sparse interval jacobian matrices. In this paper, we first propose a general indexed storage scheme to store sparse interval matrices We then present an iterative interval linear solver that utilizes the proposed index storage scheme It is expected that the newly proposed general interval iterative sparse linear solver will improve the overall performance for interval Newton/Generalized bisection methods when the jacobian matrices are sparse. In section 1, we briefly review interval Newton's methods. In Section 2, we review some currently used storage schemes for sparse systems. In Section 3, we introduce a new index scheme to store general sparse matrices. In Section 4, we present both sequential and parallel algorithms to evaluate a general sparse Jacobian matrix. In Section 5, we present both sequential and parallel algorithms to solve the corresponding interval linear system by the all-row preconditioned scheme. Conclusions and future work are discussed in Section 6.     

Parallel methods for optimality criteria-based topology optimization

Topology optimization problems require the repeated solution of finite element problems that are often extremely ill-conditioned due to highly heterogeneous material distributions. This makes the use of iterative linear solvers inefficient unless appropriate preconditioning is used. Even then, the solution time for topology optimization problems is typically very high. These problems are addressed by considering the use of non-overlapping domain decomposition-based parallel methods for the solution of topology optimization problems. The parallel algorithms presented here are based on the solid isotropic material with penalization (SIMP) formulation of the topology optimization problem and use the optimality criteria method for iterative optimization. We consider three parallel linear solvers to solve the equilibrium problem at each step of the iterative optimization procedure. These include two preconditioned conjugate gradient (PCG) methods: one using a diagonal preconditioner and one using an incomplete LU factorization preconditioner with a drop tolerance. A third substructuring solver that employs a hybrid of direct and iterative (PCG) techniques is also studied. This solver is found to be the most effective of the three solvers studied, both in terms of parallel efficiency and in terms of its ability to mitigate the effects of ill-conditioning. In addition to examining parallel linear solvers, we consider the parallelization of the iterative optimality criteria method. To tackle checkerboarding and mesh dependence, we propose a multi-pass filtering technique that limits the number of “ghost” elements that need to be exchanged across interprocessor boundaries.     

一种有效的新预条件Gauss-Seidel迭代法

为了改善古典迭代法的收敛速度,本文提出一种带参数的新预条件方法,并对参数的选择给出必要条件,证明了对于非奇异不可约M一矩阵,新预条件方法收敛且可以加速Gauss—Seidel迭代法的收敛速度,数值例子表明新预条件方法是有效的．     

A parallel preconditioned conjugate gradient solution method for finite element problems with coarse–fine mesh formulation

The object of this paper is a parallel preconditioned conjugate gradient iterative solver for finite element problems with coarse-mesh/fine-mesh formulation. An efficient preconditioner is easily derived from the multigrid stiffness matrix. The method has been implemented, for the sake of comparison, both on a IBM-RISC590 and on a Quadrics-QH1, a massive parallel SIMD machine with 128 processors. Examples of solutions of simple linear elastic problems on rectangular grids are presented and convergence and parallel performance are discussed.     

A numerical study of a 3D bioheat transfer problem with different spatial heating

We develop numerical methods for the computer simulation and modeling of a three dimensional heat transfer problem in biological bodies. The technique is intended for the temperature predications and parameter measurements in thermal medical practices and for the studies of thermomechanical interaction of biological bodies at high temperature.We examine a mathematical model based on the classical well-known Pennes equation for heat transfer in biological bodies. A finite difference discretization scheme is used to discretize the governing partial differential equation. A preconditioned iterative solver is employed to solve the resulting sparse linear system at each time step. Numerical results are obtained to demonstrate the efficacy of the proposed numerical methods.     

Implicit multigrid schemes for challenging aerodynamic simulations on block-structured grids

In this paper, the use of implicit multigrid smoothers for challenging aerodynamic simulations is explored. The block lower–upper symmetric Gauss–Seidel (LU-SGS) and hybrid Runge–Kutta/LU-SGS schemes are implemented in Bombardier’s multiblock Navier–Stokes solver, FANSC. The schemes are compared to the existing Runge–Kutta and point-Jacobi preconditioned explicit multistage smoothers. Through tests ranging from 2D airfoils to 3D wing-body-engine cases, the computational speed-up and robustness of the implicit schemes are evaluated. It is shown that the implicit smoothers present a computational speed-up of at least two, and are significantly more robust, especially for flow problems involving the “power-on” engine boundary condition.     

Performance of iterative methods in ANSYS on cray parallel/vector supercomputers

This paper describes recent work using iterative methods for the solution of linear systems in the ANSYS program. The ANSYS program, a general purpose finite element code widely used in structural analysis applications, has now added an iterative solver option. The development of robust iterative solvers and their use in commercial programs is discussed. Discussion of the applicability of iterative solvers as a general purpose solver will include the topics of robustness; as well as memory requirements and CPU performance. A new iterative solver for general purpose finite element codes which functions as a “black-box” solver using element-specific information and the underlying problem physics to construct an effective and inexpensive preconditioner is described. Some results are given from realistic examples comparing the performance of the iterative solver implemented in ANSYS with the traditional parallel/vector frontal solver used in ANSYS and a robust shifted incomplete Choleski iterative solver.     

Recycling Krylov subspaces for efficient large-scale electrical impedance tomography

Electrical impedance tomography (EIT) captures images of internal features of a body. Electrodes are attached to the boundary of the body, low intensity alternating currents are applied, and the resulting electric potentials are measured. Then, based on the measurements, an estimation algorithm obtains the three-dimensional internal admittivity distribution that corresponds to the image. One of the main goals of medical EIT is to achieve high resolution and an accurate result at low computational cost. However, when the finite element method (FEM) is employed and the corresponding mesh is refined to increase resolution and accuracy, the computational cost increases substantially, especially in the estimation of absolute admittivity distributions. Therefore, we consider in this work a fast iterative solver for the forward problem, which was previously reported in the context of structural optimization. We propose several improvements to this solver to increase its performance in the EIT context. The solver is based on the recycling of approximate invariant subspaces, and it is applied to reduce the EIT computation time for a constant and high resolution finite element mesh. In addition, we consider a powerful preconditioner and provide a detailed pseudocode for the improved iterative solver. The numerical results show the effectiveness of our approach: the proposed algorithm is faster than the preconditioned conjugate gradient (CG) algorithm. The results also show that even on a standard PC without parallelization, a high mesh resolution (more than 150,000 degrees of freedom) can be used for image estimation at a relatively low computational cost.