NA Digest Sunday, June 28, 1998 Volume 98 : Issue 24

Today's Editor:
Cleve Moler
The MathWorks, Inc.

Submissions for NA Digest:

Mail to

Information about NA-NET:

Mail to

URL for the World Wide Web:

From: Jim Epperson <>
Date: Wed, 24 Jun 1998 15:28:52 -0500 (CDT)
Subject: History of Splines

Many years ago, during my graduate school days, I heard (or thought I
heard) that some of the earliest work on splines was done at General
Motors, and that the motivation for it was to accurately model the curved
surfaces of sheet metal that formed the shell of the car bodies. Is this
actually the case? Can someone point me to a reference that could
provide more details, and actually be used as a citation to support this

Jim Epperson


From: Jeff Saltzman <>
Date: Wed, 24 Jun 1998 12:14:45 -0400
Subject: Change of Address for Jeffrey Saltzman

I am now located at the following address:

Jeffrey Saltzman
Merck and Company, Inc.
P.O. Box 2000
Rahway, NJ 07065-0900

My new e-mail address is:

E-mail will be forwarded from my address @ LANL but probably not

Jeff Saltzman


From: Eduardo Sontag <>
Date: Fri, 26 Jun 1998 00:54:49 -0400 (EDT)
Subject: New Book, Mathematical Control Theory

Announcing a new book:
Eduardo D. Sontag
Mathematical Control Theory: Deterministic Finite Dimensional Systems
***Second Edition***

Springer-Verlag, New York, 1998, ISBN 0-387-984895
May be ordered from 1-800-Springer toll-free in the USA, or via email from:; or faxing +1.201.345.4505.

This textbook introduces the core concepts and results of Control and System
Theory. Unique in its emphasis on foundational aspects, it takes a "hybrid"
approach in which basic results are derived for discrete and continuous time
scales, and discrete and continuous state variables.
Primarily geared towards mathematically advanced undergraduate or graduate
students, it may also be suitable for a second engineering course in control
which goes beyond the classical frequency domain and state-space material.
The choice of topics, together with detailed end-of-chapter links to the
bibliography, makes it an excellent research reference as well.

The Second Edition constitutes a substantial revision and extension of the
First Edition, mainly adding or expanding upon advanced material, including:
Lie-algebraic accessibility theory, feedback linearization,
controllability of neural networks, reachability under input constraints,
topics in nonlinear feedback design (such as backstepping, damping,
control-Lyapunov functions, and topological obstructions to stabilization),
and introductions to the calculus of variations, the maximum principle,
numerical optimal control, and linear time-optimal control.

Also covered, as in the First Edition, are notions of systems and
automata theory, and the algebraic theory of linear systems, including
controllability, observability, feedback equivalence, and minimality;
stability via Lyapunov, as well as input/output methods;
linear-quadratic optimal control; observers and dynamic feedback;
Kalman filtering via deterministic optimal observation; parametrization of
stabilizing controllers, and facts about frequency domain such as the
Nyquist criterion.

Chapter Headings:

Reachability and Controllability
Nonlinear Controllability
Feedback and Stabilization
Observers and Dynamic Feedback
Optimality: Value Function
Optimality: Multipliers
Optimality: Minimum-Time for Linear Systems
Appendix: Linear Algebra
Appendix: Differentials
Appendix: Ordinary Differential Equations
List of Symbols


From: Maxim Larin <>
Date: Thu, 25 Jun 98 18:43:03 +0700
Subject: New Generation Solution Methods for PDEs

International Conference on
``New Generation Solution Methods for PDE''

organized by
Sergei V. Nepomnyaschikh and Panayot S. Vassilevski

The Institute of Computational Mathematics and Mathematical
Geophysics (former Computing Center) of the Siberian Devision
of Russian Academy of Sciences (SD RAS),
in cooperation with
the Central Laboratory on Parallel Processing,
Bulgarian Academy of Sciences (BAS)
will organize international conference on
``New Generation Solution Methods for PDEs''

The purpose of the conference is to bring together leading experts in
scientific computing that are actively involved in theory and computation;
that are currently developing or can develop highly efficient ``new
generation'' solution methods for PDEs. In particular, the topics of the
conference will include methods for solving PDEs and integral equations
that come from fluid flow modeling, contaminant transport, computational
electromagnetics, flow in porous media, perfectly matched layer methods
for wave problems; and generally, from engineering practice; both stationary
and time dependent, including non--linear as well as eigen--problems and
generally parameter dependent problems. Their impact on theoretical results
as well as on solving practical problems will be the main objective of the

The conference aims at generating creative new directions and ideas for
research that can be a basis for solid collaboration, especially for the
new generation of scientists. This will help towards building a community
of young scientists with balanced participation from Eastern, Western Europe
and U.S.A. to be involved in a stable cooperation in the forthcoming years,
using the experience of well known leaders of the international scientific


The workshop is scheduled for January 11--January 16, 1999 with
arrival: January 10 and departure: noon January 16, 1999.


Institute of Computational Mathematics and Mathematical
Geophysics, Novosibirsk, Russian Academy of Sciences


Specific topics of interest are:
- Analysis of numerical methods;
- Numerical Linear Algebra;
- Basic iterative methods and robust preconditioning;
- MG/DD/FS (multigrid, domain decomposition, and
fictitious space) methods;
- Multiscale methods; theory and computation;
- Conservative schemes and mixed finite element computations;
- Adaptive methods;
- Modeling of contaminant transport and related applications;
- Flow computations;
- Computational electromagnetics (numerical solution of
Maxwell's equations);
- Perfectly matched layer methods for wave problems;
- Parallel performance;

Owe Axelsson, KUN, Nijmegen, The Netherlands;
Nikolai~S. Bakhvalov, Moscow University, Moscow, Russia;
Randolph Bank, UC at San Diego, La Jolla, U.S.A.;
Tony Chan, UC at Los Angeles, Los Angeles, U.S.A.;
Wolfgang Dahmen, RWTH, Aachen, Germany;
Craig Douglas, University of Kentucky, U.S.A.;
Roland Freund, Bell Lab \& Lucent Technology, U.S.A.;
Michael Griebel, University of Bonn, Germany;
Wolfgang Hackbusch, Kiel, Germany;
Mikhail Kaschiev, IMI--BAS, Sofia, Bulgaria;
Yuri A. Kuznetsov, RAS, Moscow, Russia/Houston, TX, U.S.A.;
Ulrich Langer, Johannes Kepler Universit{\" at}, Linz, Austria;
Raytcho Lazarov, Texas A \& M University, College Station, TX, U.S.A.;
Patrick Le Tallec, INRIA, France;
Peter Oswald, Bell Lab \& Lucent Technology, U.S.A.;
Joseph Pasciak, Texas A \& M University, College Station, TX, U.S.A.;
Jacques Periaux, Dassault Aviation, Paris, France;
Vladimir V. Shaidurov,Computing Center SD RAS, Krasnoyarsk, Russia;
Junping Wang, University of Wyoming, Laramie, WY, U.S.A.;
Jinchao Xu, Penn State University, U.S.A.;
Harry Yserentant, University of T{\" u}bingen, Germany;


Ms. Sylvia Dimitrova
Address: CLPP-BAS, ``Acad. G. Bontchev'' street, Block 25A,
BG-1113 Sofia, Bulgaria;
FAX: (--359--2)--707--273;

Dr. Maxim Larin
Address: ICMMG SD RAS, avenue Lavrentiev, 6,
630090 Novosibirsk, Russia;
FAX: (7--383--2)--343--783;


From: Bette Byrne <>
Date: Fri, 26 Jun 98 09:51:31
Subject: Workshop on Application of Adjoint Methods

One-Day Workshop on Application of Adjoint Methods
Tuesday, 21st July, 1998
University of Oxford


This one-day workshop will discuss the way in which adjoint methods play a role in design optimisation, a
posteriori error estimation, history matching and data assimilation.

The meeting affords an opportunity for researchers in different communities to pool experiences from a number of
relevant areas.

The workshop will take place at Oxford University Computing Laboratory, Wolfson Building, Parks Road, Oxford,
and is planned to begin at 10.30 a.m. with coffee.

Speakers will include:

Prof M B Giles (Oxford)
An Introduction to the Adjoint Approach to Design, Data
Assimilation and Error Analysis.

Prof N. K. Nichols (Reading)
To be arranged

Mr A Lawless (The Met. Office)
Development of adjoint models for data assimilation in

Dr P Childs, Dr J Bulman and Dr C Farmer (GeoQuest, Abingdon)
History Matching in Oil Reservoir Simulation.

Dr P Houston (Oxford)
A Posteriori Error Analysis of Stabilised Finite Element
Approximations of Hyperbolic Problems.

Dr A K Parrott (Greenwich)
Application of Adjoint Methods for Electromagnetic Design

Dr N Pierce (Oxford)
Superconvergence for the Quasi-1D Euler Equations.

If you require any further information please contact:

Mrs B C Byrne
Oxford University Computing Laboratory,
Wolfson Building, Parks Road,
Oxford OX1 3QD.
Tel: (01865) 273883
Fax: (01865) 273839


From: Marcin Paprzycki <>
Date: Sun, 28 Jun 1998 00:07:00 -0500 (CDT)
Subject: Southern Symposium on Computation


December 4-5, 1998
University of Southern Mississippi
Hattiesburg, Mississippi

Organized by:
School of Mathematical Sciences of the University of Southern Mississippi
in cooperation with the Department of Mathematics and the Department of
Statistics and Computer Science, Mississippi State University, Department
of Computer Science, Louisiana State University

The aim of the conference is to bring together researchers in all areas of
computation and in an informal atmosphere attempt to develop links between
its various threads. The emphasis is on developing contacts and links with
the the breadth of computational science, linking more closely
computational mathematics and computer science.

KEYNOTE ADDRESSESS (preliminary list):

G. Fairweather (Colorado School of Mines) - tenative
S. Iengar (Louisiana State University)
P. Lax (Courant Institute) - tenative
S. Piacsek (Stennis Space Center) - tenative
T. Skjellum (Mississippi State University)
L. Welch (UT Arlington) - tenative
J. Zhu (Mississippi State University)


J. Banicescu (Mississippi State University, USA)
I. Bar-On (Technion University, Israel)
A. Blazniewskij (University of Podillia, Ukraine)
C. Breshears (Nichols Research, Vicksburg, USA)
I. Gladwell (Southern Methodist University, USA)
J. Leszczynski (Jackson State University, USA)
X. Li (Loyola University, New Orleans, USA)
N. Mastorakis (Hellenic Naval Academy, Greece)
F. Mazzia (University of Bari, Italy)
I. Sivergina (Jekaterinburg Tech, Russia)
P. Stpiczynski (UMCS, Poland)
T. Taha (University of Georgia, USA)
J. Taylor (Louisiana State University, USA)
P. Yalamov (University of Rousse, Bulgaria)
W. Zhang (Louisiana Tech, USA)
B. Zubik-Kowal (University of Leiden, Netherlands)

Organizing Committee:
M. Paprzycki (Chair)
M. Cobb, J. Ding, J. Kolibal, M. Mascagni, R. Necaise, R. Pandey,
W. Russell.

We invite contributions covering all areas of computation. Special
sessions are also welcome. Contributions from graduate students are also


Special sesion proposals due: September 30, 1998
Extended abstracts (1 page) due: October 31, 1998
Acceptence: November 1, 1998
Papers for proceedings: November 31, 1998

Electronic proceedings will be published.

Information about the meeting will be updated on the conference WWW site
to be located at:
To obtain more information about the meeitng send e-mail to:


From: Dominique Chapelle <>
Date: Fri, 26 Jun 1998 20:56:00 +0200
Subject: Faculty Position at Pole Universitaire Leonard de Vinci

Position of Assistant or Associate Professor in Applied
Mathematics/Scientific Computing.

The Pole Universitaire Leonard de Vinci (Paris-la-Defense) welcomes
applications for a professorship in applied mathematics, with a
specialization in Scientific Computing, at the Assistant or Associate
(preferably) Professor level. This position is to be filled as soon as

Candidates should meet the following requirements:

a) Hold a PhD (or equivalent degree) in applied mathematics (numerical
analysis or scientific computing).

b) Have an excellent practice of scientific computing.

They are expected to:

a) Teach (mostly in French) in mathematics at various
undergraduate/graduate levels.

b) Perform research within the MOSTRA research group (operating jointly
with INRIA-Rocquencourt) on the following topics:
- design and analysis of efficient numerical procedures for structural
models ;
- solvers for large-scale computations in solid mechanics ;
- numerical simulation of fluid-structure interaction problems ;
- adaptive structures.
Preference will be given to candidates with research
interests/background oriented towards active control of structures.

c) Develop industrial collaborations.

Applications, including a detailed CV and a list of three referees
(with name and title, surface and email addresses, phone and fax
numbers), should be sent to:

Pr. Michel Bernadou
DER Calcul Scientifique
Pole Universitaire Leonard de Vinci
92916 Paris-la-Defense Cedex

The Pole Universitaire Leonard de Vinci, a new private university, is
located in Paris-la-Defense. Faculty members teach and perform research
in fields such as Applied Mathematics, Mechanics, Computer Science,
Marketing and Management.

Inquiries regarding the joint INRIA/PULV research project (MOSTRA) can
be directed to


From: David W Walker <>
Date: Wed, 24 Jun 1998 15:44:42 +0100
Subject: Research Positions at Cardiff University

Advanced Interfaces and Tools for Distributed Scientific Computing

Postdoctoral and PhD student positions
Department of Computer Science
Cardiff University

The Parallel and Scientific Computing Group in the Department of Computer
Science at Cardiff University is seeking qualified individuals for an
EPSRC-funded research project (GR/M17327) entitled "Problem-Solving
Environments for Large-Scale Simulation." A two-year Post-Doctoral Research
Associate position, and a three-year postgraduate studentship leading towards
a PhD, are available. Further details are available at:

The intended start date for the positions is 1st October 1998.


From: Maya Neytcheva <>
Date: Wed, 24 Jun 1998 08:54:23 +0200 (MET DST)
Subject: Contents, Numerical Linear Algebra with Applications

Numerical Linear Algebra with Applications
Volume 5, Issue 1, 1998

A note on the normwise perturbation theory for the regular
generalized eigenproblem
V. Fraysse and V. Toumazou

A parallel multisplitting solution of the Least Squares problem
R.A. Renaut

Harmonic projection methods for large nonsymmetric eigenvalue problems
R. Morgan and Min Zeng

Efficient computation of sparse approximate inverses
T.K. Huckle

Numerical Linear Algebra with Applications
Volume 5, Issue 2, 1998

Approximating minimum norm solutions of rank-deficient Least Squares
A. Dax and L. Elden

On the performance of various adaptive preconditioned
GMRES strategies
K. Burrage and J. Erhel

On parallel solution of linear elasticity problems. Part I: Theory
I. Gustafson and G. Lindskog

Numerical Linear Algebra with Applications
Volume 5, Issue 3, 1998

Estimating the largest eigenvector by Lanczos and polynomial algorithms
with a random start
Z. Leyk and H. Wozniakowski

Analysis of matrix-dependent multigrid algorithms
Y. Shapira

The structure of the boundary element matrix for the three-dimensional
Dirichlet problem in elasticity
S. Rjasanow

Indefinitely preconditioned inexact Newton method for large sparse
equality constrained nonlinear programming problems
L. Lucsan and J. Vlcek


From: Elsevier Science <>
Date: Thu, 25 Jun 1998 22:41:48 +0100 (BST)
Subject: Contents, Computer Physics Communications

ISSN : 0010-4655
Vol./Iss. : 110 / 1-3

Computing at KEK
Morita , Y.
pp.: 1-5

Computing trends and strategy at CERN
Robertson , Les
pp.: 6-11

Computing at DESY -- current setup, trends and strategic
Ernst , Michael
pp.: 12-17

Strategic directions of computing at Fermilab
Wolbers , Stephen
pp.: 18-21

Belle computing model
Katayama , Nobu
pp.: 22-25

ATLAS computing
Knobloch , Ju"rgen
pp.: 26-31

Computing models of CDF and DO in Run II
Lammel , Stephan
pp.: 32-37

The BaBar computing model
Geddes , Neil
pp.: 38-42

Networking, videoconferencing and collaborative environments
Galvez , P.
pp.: 43-50

Collaborative computing environments for HEP
Bunn , Julian J.
pp.: 51-58

Multiprocessor design options and the Silicon Graphics S2MP
Lenoski , Daniel
pp.: 59-68

Trends and challenges in data acquisition and control
systemsSummary of CHEP97 Session B
Lankford , A.J.
pp.: 69-75

Large systems and specific solutionsSummary of CHEP97
Session F
Gagliardi , Fabrizio
pp.: 76-79

On the treatment of energy loss in track fitting
Fru"hwirth , R.
pp.: 80-86

SMI++ -- Object-oriented framework for designing control
systems for HEP experiments
Gaspar , C.
pp.: 87-90

The DAQ system for CLEO III
Wolf , A.
pp.: 91-94

The ATLAS DAQ and Event Filter prototype ``-1" project
Ambrosini , G.
pp.: 95-102

CMS online event filter software
Kruse , Andre's
pp.: 103-106

PHENIX on-line distributed computing system architecture
Desmond , Edmond
pp.: 107-112

Software technologies for a prototype ATLAS DAQ
Burckhart , D.
pp.: 113-119

An ODMG-compatible testbed architecture for scalable
management and analysis of physics data
Malon , David M.
pp.: 120-124

ODBMS for LHC detector simulation
Asai , Makoto
pp.: 125-130

A distributed disk layer for mass storage at DESY
Brand , Susanne
pp.: 131-133

The reality of collaboratories
Agarwal , D.A.
pp.: 134-141

What is the Internet doing? Performance and reliability
monitoring for the HEP community
Cottrell , R.L.A.
pp.: 142-148

The multi-session bridge
Roediger , G.A.
pp.: 149-154

WIRED -- World-Wide Web Interactive Remote Event Display
Coperchio , M.C.
pp.: 155-159

The Moose project
Bos , Kors
pp.: 160-163

Analysis, design, and implementation of PHENIX on-line
computing systems software using Shlaer--Mellor
Object-Oriented Analyis and Recursive Design
Kozlowski , Thomas
pp.: 164-169

Workflow management in the assembly of CMS ECAL
Baker , N.
pp.: 170-176

MISCOMP -- an information system tool-kit
White , Vicky
pp.: 177-180

A WWW software development environment to support
cooperative and spread working groups
Maidantchik , C.
pp.: 181-191

CMS reconstruction and analysis: an object oriented approach
Innocente , Vincenzo
pp.: 192-197

A taxonomically distributed data retrieval model
Marquina , M.
pp.: 198-205

Results from the Macrame' 1024 node switching network
Haas , S.
pp.: 206-210

Next generation farms at Fermilab
Cudzewicz , Ron
pp.: 211-215

The teraflop supercomputer APEmille: architecture, software
and project status report
Aglietti , F.
pp.: 216-219

The CP-PACS parallel computer
Ukawa , Akira
pp.: 220-224

An evaluation of PCs for high energy physics under Windows
NT and Linux
Athanas , Michael
pp.: 225-229

Supporting Windows NT as physics workstations and servers
Foster , David
pp.: 230-238


End of NA Digest