Riccati equation for infinite-dimensional discrete bilinear systems

The problem of quadratic optimal feedback control for infinite-dimensionaldiscrete bilinear systems is associated with an algebraic Riccati-likeoperator equation. The existence and uniqueness of a nonnegativesolution to such a Riccati-like operator equation is establishedas the strong limit of a sequence of solutions to Lyapunov-likeoperator equations. This research was supported in part by CNPq (Brazilian NationalResearch Council) and FAPESP (Research Council of the Stateof sào Paulo).     

Mixed block elimination for linear systems with wider borders

The paper is about the stable solution of possibly ill-conditionedbordered linear systems. Given stable solvers for matrix A andfor A^T, we prove that the Govaerts Mixed Block Elimination (BEM)method constitutes a stable solver for the matrix consistingof A or A^T with a border of width 1, and hence by recursionfor a border of any width. We express the algorithm in an efficient,iterative, form. We analyse its operation count, and verifythe theory by extensive numerical experiments. ^*Senior Research Associate of the Belgian National Fund of ScientificResearch NFWO.     

A new pseudospectral method with upwind features

A new pseudospectral method is presented for numerical solutionsof singular perturbation problems without turning points. Boththeoretical and numerical analyses show that this new methodis an upwind scheme. It is shown that when the perturbationparameter is fixed the computed solution converges spectrallyto the exact solution as the number of collocation points tendsto infinity. ^* Supported by a Royal Fellowship Award from the Royal Societyof London.     

The Caratheodory--Fejer Extension of a Finite Geometric Series

It is shown that the Caratheodory—Fejer extension of afinite geometric series can be given explicitly up to a simplepolynomial equation in an auxiliary variable. This result allowsus to analyse the Caratheodory-Fejer approximation method inthe case where the quotients of successive Maclaurin coefficientsof the given function tend to a limit. ^*Research carried out while this author was at ETH Zurich partiallysupported by a Royal Society European Visiting Fellowship.     

The Polynomial Caratheodory--Fejer Approximation Method for Jordan Regions

We propose a method for the approximation of analytic functionson Jordan regions that is based on a Carathéodory—Fejértype of economization of the Faber series. The method turnsout to be very effective if the boundary of the region is analytic.It often still works when the region degenerates to a Jordanarc. We also derive related lower and upper bounds for the errorof the best approximation. ^*Research carried out while this author was at ETH Zurich partiallysupported by a Royal Society European Visiting Fellowship.     

Newton additive and multiplicative Schwarz iterative methods

Daniel B. Szyld^¶ Department of Mathematics, Temple University, Philadelphia, PA 19122, USA ^{Convergence properties are presented for Newton additive andmultiplicative Schwarz (AS and MS) iterative methods for thesolution of nonlinear systems in several variables. These methodsconsist of approximate solutions of the linear Newton step usingeither AS or MS iterations, where overlap between subdomainscan be used. Restricted versions of these methods are also considered.These Schwarz methods can also be used to precondition a Krylovsubspace method for the solution of the linear Newton steps.Numerical experiments on parallel computers are presented, indicatingthe effectiveness of these methods.}     

Adaptive frame methods for elliptic operator equations: the steepest descent approach

Massimo Fornasier^¶ Dipartimento di Metodi e Modelli Matematici per le Scienze Applicate, Università "La Sapienza" in Roma, Via Antonio Scarpa, 16/B, I-00161 Roma, Italy Rob Stevenson^|| Department of Mathematics, Utrecht University, PO Box 80.010, NL-3508 TA Utrecht, The Netherlands ^{This paper is concerned with the development of adaptive numericalmethods for elliptic operator equations. We are particularlyinterested in discretization schemes based on wavelet frames.We show that by using three basic subroutines an implementable,convergent scheme can be derived, which, moreover, has optimalcomputational complexity. The scheme is based on adaptive steepestdescent iterations. We illustrate our findings by numericalresults for the computation of solutions of the Poisson equationwith limited Sobolev smoothness on intervals in 1D and L-shapeddomains in 2D.}     

A discrete-time reduced-order controller for robust stabilization of plant in the normalized coprime-factor description

Using the technique of singlar perturbation, a procedure fordesigning a discrete–time reduced–order controllerfor robust stabilization is proposed. The theoretical aspectsare derived directly from the results of the generalized Popov–Yakubovichtheory recently developed. ¹Vlad Ionescu, 3 Emile Zola, Bucharest, 71272, Romania. ²Cristian Oara, 34 Austrului, Bucharest, 73115, Romania.     

A note on interval-polynomial stability

In this note, a theorem of Kharitonov and a theorem of XieNare combined to obtain an approach which simplifies dealingwith the problems of interval–polynomial stability. ^*On leave with Sibley School of Mechanical & Aerospace Engineering,Cornell Univ., Ithaca, NY 14853, USA.     

Maximum Principles for Optimal Controls for Elliptic Variational Inequalities of the Second Kind

A maximum principle is obtained for optimal controls for someelliptic variational inequalities of the second kind by usingthe penalty method, Ekelans's variational principle, and lowersemicontinuity of some set-valued mappings. It has been shownthat this principle leads to some known optimality conditionsin many cases. It also yields new optimality conditions. Insome cases, it leads to an analogy of pontryagin's principle. Permanent address: Mathematics Department, Centre South Universityof Technology, China.     

Geometrically Derived Difference Formulae for the Numerical Integration of Trajectory Problems

The term "trajectory problem" is taken to include problems thatcan arise, for instance, in connection with contour plotting,or in the application of continuation methods, or during phaseplane analysis. Geometrical techniques are used to constructdifference methods for these problems to produce in turn explicitand implicit circularly exact formulae. Based on these formulae,a predictor-corrector method is derived which, when comparedwith a closely related standard method, shows improved performance.It is found that this latter method produces spurious limitcycles, and this behaviour is partly analysed. Finally, a simplevariable-step algorithm is constructed and tested. Visting Scientists at the National Research Institute for MathematicalSciences, CSIR, P.O. BOX 395. pretoria, South Africa     

The multidimensional positive and nearly conservative Lagrange-Galerkin method. Part II: the use of C1 elements

In this paper we consider a monotonic and conservative variationon the direct Lagrange-Galerkin method due to Priestley, andaddress certain problems with both the original Lagrange-Galerkinmethod and this new method which disappear with the use of higher-orderC¹ basis functions. Here we concentrate on the use of Hermite-cubics,demonstrating their superior stability in the Lagrange- Galerkinmethod, and the adaptations of the recovery techniques neededto ensure monotonicity and near conservation. Current address: Schlumberger GeoQuest, Abingdon Business Park,Oxon OX14 10 UK.     

Numerical treatment of retarded boundary integral equations by sparse panel clustering

Stefan Sauter Institute for Mathematics, University of Zürich, Winterthurerstrasse 190, CH-8057 Zürich, Switzerland ^{We consider the wave equation in a boundary integral formulation.The discretization in time is done by using convolution quadraturetechniques and a Galerkin boundary element method for the spatialdiscretization. In a previous paper, we have introduced a sparseapproximation of the system matrix by cut-off, in order to reducethe storage costs. In this paper, we extend this approach byintroducing a panel clustering method to further reduce thesecosts.}     

On the perturbation of LU and Cholesky factors

In a recent paper, Chang and Paige have shown that the usualperturbation bounds for Cholesky factors can systematicallyoverestimate the errors. In this note we sharpen their resultsand extend them to the factors of the LU decomposition. Theresults are based on a new formula for the first-order termsof the error in the factors. ^* This report is available by anonymous ftp from thales.cs.umd.eduin the directory pub/reports.     

An analysis of the coupling of finite-element and Nystrm methods in acoustic scattering

In this paper we analyze the use of a combined Nystrm and finite-elementprocedure for approximating the scattered time-harmonic acousticfield produced when an incident wave interacts with a boundedinhomogeneity. The coupling technique uses a smooth artificialboundary and explicitly decouples the integral equation andfinite-element computations. We prove convergence of the method.One highlight of this analysis is that we prove Sobolev spaceerror estimates for the Nystrm scheme (which is usually analyzedin Hlder spaces). We also present some numerical results. E-mail: kirsch{at}am.uni-erlangen.de E-mail: monk{at}math.udel.edu     

Algebraic Manipulation of Polynomial Matrices: A Survey

A concise review is given of some recent methods for performingalgebraic operations on polynominal matrices. The problems addressedare division, inversion, and extraction of greatest common divisorof a pair of matrices. Based on a presentation at the International Symposium on theMathematical theory of Networks and Systems, Amsterdam, June1989.     

Qualitative dynamics, root-loci and asymptotic structure of universal adaptive stabilizers

A problem of universal adaptive stabilization in Rⁿ is approachedin a qualitative manner to relate the form of the trajectoriesin the extended state space Rⁿ⁺¹ to the root locus of the associatedfixed-parameter linear system. Relationships are derived betweenthe values of the limit gain and the initial conditions. Numericalstudies are used to illustrate and support these ideas by computationof generic trajectories and attempted computation of certainnon-generic possibilities. The implications of the study formore general dynamic situations are outlined. These authors are also with the Department of Mathematics,University of Exeter, EX4 4QE.     

Estimates of the region of attraction of continuous-time cascade systems

Xiao-Song Yang Department of Mathematics, Huazhong University of Science and Technology Wuhan, Hubei 430074, People's Republic of China Huimin Li Departments of Mathematics and Control and Engineering, Huazhong University of Science and Technology Wuhan, Hubei 430074, People's Republic of China Corresponding author. Email: yangxs{at}cqupt.edu.cn Received on May 1, 2006; Accepted on October 6, 2006 In this paper, we first present some new sufficient conditionsfor global asymptotic stability of continuous-time cascade systems.Then, we give some detailed results on the region of attractionof continuous-time cascade systems. Some examples are presentedto illustrate these results.     

On the Stability Analysis of Nonlinear Systems

The stability of nonautonomous delay differential equationsis studied via a combination of the global linerization andnonlinear variation of parameter techniques. Permanent address: Department of Mathematics, University ofElectronics Science & Technology of China, Chengdu, China.     

Sparse convolution quadrature for time domain boundary integral formulations of the wave equation

S. A. Sauter Institut für Mathematik, Universität Zürich, Winterthurerstrasse 190, CH-8057 Zürich, Switzerland ^{Many important physical applications are governed by the waveequation. The formulation as time domain boundary integral equationsinvolves retarded potentials. For the numerical solution ofthis problem, we employ the convolution quadrature method forthe discretization in time and the Galerkin boundary elementmethod for the space discretization. We introduce a simple apriori cut-off strategy where small entries of the system matricesare replaced by zero. The threshold for the cut-off is determinedby an a priori analysis which will be developed in this paper.This analysis will also allow to estimate the effect of additionalperturbations such as panel clustering and numerical integrationon the overall discretization error. This method reduces thestorage complexity for time domain integral equations from O(M2N)to O(M2N logM), where N denotes the number of time steps andM is the dimension of the boundary element space.}