General theory of domain decomposition: Indirect methods

According to a general theory of domain decomposition methods (DDM), recently proposed by Herrera, DDM may be classified into two broad categories: direct and indirect (or Trefftz‐Herrera methods). This article is devoted to formulate systematically indirect methods and apply them to differential equations in several dimensions. They have interest since they subsume some of the best‐known formulations of domain decomposition methods, such as those based on the application of Steklov‐Poincaré operators. Trefftz‐Herrera approach is based on a special kind of Green's formulas applicable to discontinuous functions, and one of their essential features is the use of weighting functions which yield information, about the sought solution, at the internal boundary of the domain decomposition exclusively. A special class of Sobolev spaces is introduced in which boundary value problems with prescribed jumps at the internal boundary are formulated. Green's formulas applicable in such Sobolev spaces, which contain discontinuous functions, are established and from them the general framework for indirect methods is derived. Guidelines for the construction of the special kind of test functions are then supplied and, as an illustration, the method is applied to elliptic problems in several dimensions. A nonstandard method of collocation is derived in this manner, which possesses significant advantages over more standard procedures. © 2002 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq 18: 296–322, 2002; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/num.10008     

A general effective method for combining collocation and DDM: An application of discontinuous Galerkin methods

Domain decomposition methods (DDM) have received much attention in recent years. They constitute the most effective means of using parallel computing resources to model continuous systems. However, combining collocation procedures with domain decomposition methods presents complications that must be overcome in order to profit from the advantages of parallel computing. The present paper belongs to a line of research in which a theory that constitutes a general and systematic formulation of discontinuous Galerkin methods (dG) is being investigated. Based on it, a new method of collocation of general applicability, TH‐collocation, was recently introduced. For a broad class of symmetric and positive continuous systems, TH‐collocation yields symmetric and positive matrices. This clears the way for applying effectively DDM and parallel computing, in combination with collocation, to such systems. In this paper the general procedure is explained with some detail and then is applied to develop an effective method for processing elliptic equations of second order. This, by the way, overcomes the difficulties encountered in a previous Herrera and Pinder's article. © 2004 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq, 2005.     

Theory of differential equations in discontinuous piecewise‐defined functions

A truly general and systematic theory of finite element methods (FEM) should be formulated using, as trial and test functions, piecewise‐defined functions that can be fully discontinuous across the internal boundary, which separates the elements from each other. Some of the most relevant work addressing such formulations is contained in the literature on discontinuous Galerkin (dG) methods and on Trefftz methods. However, the formulations of partial differential equations in discontinuous functions used in both of those fields are indirect approaches, which are based on the use of Lagrange multipliers and mixed methods, in the case of dG methods, and the frame, in the case of Trefftz method. This article addresses this problem from a different point of view and proposes a theory, formulated in discontinuous piecewise‐defined functions, which is direct and systematic, and furthermore it avoids the use of Lagrange multipliers or a frame, while mixed methods are incorporated as particular cases of more general results implied by the theory. When boundary value problems are formulated in discontinuous functions, well‐posed problems are boundary value problems with prescribed jumps (BVPJ), in which the boundary conditions are complemented by suitable jump conditions to be satisfied across the internal boundary of the domain‐partition. One result that is presented in this article shows that for elliptic equations of order 2m, with m ≥ 1, the problem of establishing conditions for existence of solution for the BVPJ reduces to that of the “standard boundary value problem,” without jumps, which has been extensively studied. Actually, this result is an illustration of a more general one that shows that the same happens for any differential equation, or system of such equations that is linear, independently of its type and with possibly discontinuous coefficients. This generality is achieved by means of an algebraic framework previously developed by the author and his collaborators. A fundamental ingredient of this algebraic formulation is a kind of Green's formulas that simplify many problems (some times referred to as Green‐Herrera formulas). An important practical implication of our approach is worth mentioning: “avoiding the introduction of the Lagrange multipliers, or the ‘frame’ in the case of Trefftz‐methods, significantly reduces the number of degrees of freedom to be dealt with.” © 2006 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq, 2007     

Domain decomposition for radial basis meshless methods

Both overlapping and nonoverlapping domain decomposition methods (DDM) on matching and nonmatching grid have been developed to couple with the meshless radial basis function (RBF) method. Example shows that overlapping DDM with RBF can achieve much better accuracy with less nodal points compared to FDM and FEM. Numerical results also show that nonmatching grid DDM can achieve almost the same accuracy within almost the same iteration steps as the matching grid case; hence our method is very attractive, because it is much easier to generate nonmatching grid just by putting blocks of grids together (for both overlapping and nonoverlapping), where each block grid can be generated independently. Also our methods are showed to be able to handle discontinuous coefficient. © 2004 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq 20: 450–462, 2004.     

The multipliers‐free domain decomposition methods

This article concludes the development and summarizes a new approach to dual‐primal domain decomposition methods (DDM), generally referred to as “the multipliers‐free dual‐primal method.” Contrary to standard approaches, these new dual‐primal methods are formulated without recourse to Lagrange‐multipliers. In this manner, simple and unified matrix‐expressions, which include the most important dual‐primal methods that exist at present are obtained, which can be effectively applied to floating subdomains, as well. The derivation of such general matrix‐formulas is independent of the partial differential equations that originate them and of the number of dimensions of the problem. This yields robust and easy‐to‐construct computer codes. In particular, 2D codes can be easily transformed into 3D codes. The systematic use of the average and jump matrices, which are introduced in this approach as generalizations of the “average” and “jump” of a function, can be effectively applied not only at internal‐boundary‐nodes but also at edges and corners. Their use yields significant advantages because of their superior algebraic and computational properties. Furthermore, it is shown that some well‐known difficulties that occur when primal nodes are introduced are efficiently handled by the multipliers‐free dual‐primal method. The concept of the Steklov–Poincaré operator for matrices is revised by our theory and a new version of it, which has clear advantages over standard definitions, is given. Extensive numerical experiments that confirm the efficiency of the multipliers‐free dual‐primal methods are also reported here. © 2009 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq 2010     

Multipliers‐free dual‐primal domain decomposition methods for nonsymmetric matrices and their numerical testing

The most commonly used nonoverlapping domain decomposition algorithms, such as the FETI‐DP and BDDC, require the introduction of discontinuous vector spaces. Most of the works on such methods are based on approaches that originated in Lagrange multipliers formulations. Using a theory of partial differential equations formulated in discontinuous piecewise‐defined functions, introduced and developed by Herrera and his collaborators through a long time span, recently the authors have developed an approach to domain decomposition methods in which general problems with prescribed jumps are treated at the discrete level. This yields an elegant and general direct framework that permits analyzing the problems in greater detail. The algorithms derived using it have properties similar to those of well‐established methods such as FETI‐DP, but, in our experience, they are easier to implement. Also, they yield explicit matrix formulas that unify the different methods. Furthermore, this multipliers‐free framework has permitted us to extend such formulas to make them applicable to nonsymmetric matrices. The extension of the unifying matrix formulas to nonsymmetric matrices is the subject matter of the present article. A conspicuous result is that in numerical experiments in 2D and 3D, the MF‐DP algorithms for nonsymmetric matrices exhibit an efficiency of the same order as state‐of‐the‐art algorithms for symmetric matrices, such as BDDC, FETI‐DP, and MF‐DP.© 2010 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq 27: 1262‐1289, 2011     

Conservative flux recovery from the Q1 conforming finite element method on quadrilateral grids

Compared with standard Galerkin finite element methods, mixed methods for second‐order elliptic problems give readily available flux approximation, but in general at the expense of having to deal with a more complicated discrete system. This is especially true when conforming elements are involved. Hence it is advantageous to consider a direct method when finding fluxes is just a small part of the overall modeling processes. The purpose of this article is to introduce a direct method combining the standard Galerkin Q1 conforming method with a cheap local flux recovery formula. The approximate flux resides in the lowest order Raviart‐Thomas space and retains local conservation property at the cluster level. A cluster is made up of at most four quadrilaterals. © 2004 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq 20: 104–127, 2004     

Classification of positive quaternion-Kähler 12-manifoldsClassification de variétés Kähleriennes quaternioniques positives de dimension 12

We prove that the 12-dimensional complete quaternion-Kähler manifolds with positive scalar curvature belong to the list of symmetric spaces given by Wolf [12]. To cite this article: H. Herrera, R. Herrera, C. R. Acad. Sci. Paris, Ser. I 334 (2002) 43–46     

A parallel balance scheme for banded linear systems

A parallel algorithm is proposed for the solution of narrow banded non‐symmetric linear systems. The linear system is partitioned into blocks of rows with a small number of unknowns common to multiple blocks. Our technique yields a reduced system defined only on these common unknowns which can then be solved by a direct or iterative method. A projection based extension to this approach is also proposed for computing the reduced system implicitly, which gives rise to an inner–outer iteration method. In addition, the product of a vector with the reduced system matrix can be computed efficiently on a multiprocessor by concurrent projections onto subspaces of block rows. Scalable implementations of the algorithm can be devized for hierarchical parallel architectures by exploiting the two‐level parallelism inherent in the method. Our experiments indicate that the proposed algorithm is a robust and competitive alternative to existing methods, particularly for difficult problems with strong indefinite symmetric part. Copyright © 2001 John Wiley & Sons, Ltd.     

The Gauss Theorem for Domain Decompositions in Sobolev Spaces

This paper belongs to a broad line of research leaded by Herrera, which encompasses a good number of numerical methods such as Localized Adjoint Method (LAM), Eulerian-Lagrangian LAM (ELAM) and Trefftz-Herrera Method. The results presented in this paper are required in order to incorporate Herrera's general theory in a Sobolev-space setting. In particular, this article introduces a class of partitions (or domain decompositions) whose internal boundaries belong to a category of manifolds with corners, here also presented. Then a version of Gauss (or divergence) theorem, in a wider sense, is established and an explicit integral formula is associated for any given linear partial differential operator L, its adjoint and concomitant. The structure of the bilinear concomitant induced by L is first determined. Then the required formula is given over that class of domain decompositions. Finally, an integral formula well on the way of the Green-Herrera formula is settled.     

Projective synchronization between different fractional‐order hyperchaotic systems with uncertain parameters using proposed modified adaptive projective synchronization technique

In the present article, the authors have proposed a modified projective adaptive synchronization technique for fractional‐order chaotic systems. The adaptive projective synchronization controller and identification parameters law are developed on the basis of Lyapunov direct stability theory. The proposed method is successfully applied for the projective synchronization between fractional‐order hyperchaotic Lü system as drive system and fractional‐order hyperchaotic Lorenz chaotic system as response system. A comparison between the effects on synchronization time due to the presence of fractional‐order time derivatives for modified projective synchronization method and proposed modified adaptive projective synchronization technique is the key feature of the present article. Numerical simulation results, which are carried out using Adams–Boshforth–Moulton method show that the proposed technique is effective, convenient and also faster for projective synchronization of fractional‐order nonlinear dynamical systems. Copyright © 2013 John Wiley & Sons, Ltd.     

Linear system solution by null‐space approximation and projection (SNAP)

Solutions of large sparse linear systems of equations are usually obtained iteratively by constructing a smaller dimensional subspace such as a Krylov subspace. The convergence of these methods is sometimes hampered by the presence of small eigenvalues, in which case, some form of deflation can help improve convergence. The method presented in this paper enables the solution to be approximated by focusing the attention directly on the ‘small’ eigenspace (‘singular vector’ space). It is based on embedding the solution of the linear system within the eigenvalue problem (singular value problem) in order to facilitate the direct use of methods such as implicitly restarted Arnoldi or Jacobi–Davidson for the linear system solution. The proposed method, called ‘solution by null‐space approximation and projection’ (SNAP), differs from other similar approaches in that it converts the non‐homogeneous system into a homogeneous one by constructing an annihilator of the right‐hand side. The solution then lies in the null space of the resulting matrix. We examine the construction of a sequence of approximate null spaces using a Jacobi–Davidson style singular value decomposition method, called restarted SNAP‐JD, from which an approximate solution can be obtained. Relevant theory is discussed and the method is illustrated by numerical examples where SNAP is compared with both GMRES and GMRES‐IR. Copyright © 2006 John Wiley & Sons, Ltd.     

Stability and convergence of optimum spectral non‐linear Galerkin methods

Our objective in this article is to present some numerical schemes for the approximation of the 2‐D Navier–Stokes equations with periodic boundary conditions, and to study the stability and convergence of the schemes. Spatial discretization can be performed by either the spectral Galerkin method or the optimum spectral non‐linear Galerkin method; time discretization is done by the Euler scheme and a two‐step scheme. Our results show that under the same convergence rate the optimum spectral non‐linear Galerkin method is superior to the usual Galerkin methods. Finally, numerical example is provided and supports our results. Copyright © 2001 John Wiley & Sons, Ltd.     

Multipole fast algorithm for the least‐squares approach of the method of fundamental solutions for three‐dimensional harmonic problems

In this article we describe an improvement in the speed of computation for the least‐squares method of fundamental solutions (MFS) by means of Greengard and Rokhlin's FMA. Iterative solution of the linear system of equations is performed for the equations given by the least‐squares formulation of the MFS. The results of applying the method to test problems from potential theory with a number of boundary points in the order of 80,000 show that the method can achieve fast solutions for the potential and its directional derivatives. The results show little loss of accuracy and a major reduction in the memory requirements compared to the direct solution method of the least squares problem with storage of the full MFS matrix. The method can be extended to the solution of overdetermined systems of equations arising from boundary integral methods with a large number of boundary integration points. © 2003 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq 19: 828–845, 2003.     

Distributed optimal control of the viscous Burgers equation via a Legendre pseudo‐spectral approach

This paper presents a computational technique based on the pseudo‐spectral method for the solution of distributed optimal control problem for the viscous Burgers equation. By using pseudo‐spectral method, the problem is converted to a classical optimal control problem governed by a system of ordinary differential equations, which can be solved by well‐developed direct or indirect methods. For solving the resulting optimal control problem, we present an indirect method by deriving and numerically solving the first‐order optimality conditions. Numerical tests involving both unconstrained and constrained control problems are considered. Copyright © 2015 John Wiley & Sons, Ltd.     

General properties of some families of graphs defined by systems of equations

In this paper we present a simple method for constructing infinite families of graphs defined by a class of systems of equations over commutative rings. We show that the graphs in all such families possess some general properties including regularity and biregularity, existence of special vertex colorings, and existence of covering maps—hence, embedded spectra—between every two members of the same family. Another general property, recently discovered, is that nearly every graph constructed in this manner edge‐decomposes either the complete, or complete bipartite, graph which it spans. In many instances, specializations of these constructions have proved useful in various graph theory problems, but especially in many extremal problems. A short survey of the related results is included. We also show that the edge‐decomposition property allows one to improve existing lower bounds for some multicolor Ramsey numbers. © 2001 John Wiley & Sons, Inc. J Graph Theory 38: 65–86, 2001     

Nonnested multigrid methods for linear problems

This article presents an application of nonnested and unstructured multigrid methods to linear elastic problems. A variational formulation for transfer operators and some multigrid strategies, including adaptive algorithms, are presented. Expressions for the performance evaluation of multigrid strategies and its comparison with direct and preconditioned conjugate gradient algorithms are also presented. A C++ implementation of the multigrid algorithms and the quadtree and octree data structures for transfer operators are discussed. Some two‐ and three‐dimensional elasticity examples are analyzed. © 2001 John Wiley & Sons, Inc. Numer Methods Partial Differential Eq 17:313–331, 2001     

Singular Arcs in the Generalized Goddard’s Problem

We investigate variants of Goddard’s problems for nonvertical trajectories. The control is the thrust force, and the objective is to maximize a certain final cost, typically, the final mass. In this article, performing an analysis based on the Pontryagin maximum principle, we prove that optimal trajectories may involve singular arcs (along which the norm of the thrust is neither zero nor maximal), that are computed and characterized. Numerical simulations are carried out, both with direct and indirect methods, demonstrating the relevance of taking into account singular arcs in the control strategy. The indirect method that we use is based on our previous theoretical analysis and consists in combining a shooting method with an homotopic method. The homotopic approach leads to a quadratic regularization of the problem and is a way to tackle the problem of nonsmoothness of the optimal control. Support from the French Space Agency CNES (Centre National d’Etudes Spatial) is gratefully acknowledged.     

An efficient indirect RBFN‐based method for numerical solution of PDEs

This article presents an efficient indirect radial basis function network (RBFN) method for numerical solution of partial differential equations (PDEs). Previous findings showed that the RBFN method based on an integration process (IRBFN) is superior to the one based on a differentiation process (DRBFN) in terms of solution accuracy and convergence rate (Mai‐Duy and Tran‐Cong, Neural Networks 14(2) 2001, 185). However, when the problem dimensionality N is greater than 1, the size of the system of equations obtained in the former is about N times as big as that in the latter. In this article, prior conversions of the multiple spaces of network weights into the single space of function values are introduced in the IRBFN approach, thereby keeping the system matrix size small and comparable to that associated with the DRBFN approach. Furthermore, the nonlinear systems of equations obtained are solved with the use of trust region methods. The present approach yields very good results using relatively low numbers of data points. For example, in the simulation of driven cavity viscous flows, a high Reynolds number of 3200 is achieved using only 51 × 51 data points. © 2005 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq, 2005