 |
|
Publications, papers published in June 1999/communications publiées en juin 1999
- 1
-
L. Dieterle, K. Ehrenfried, R. Stuff, G. Schneider, P. Coton, J. C. Monnier,
and J.-F. Lozier.
Quantitative flow field measurements in a catapult facility using
particle image velocimetry.
In 18th International Congress on Instrumentation in Aerospace
Simulation (ICIASF), Toulouse (France), 14 - 17 June 1999.
In order to investigate the long term behaviour of the wake vortex
flow behind an aircraft, quantitative whole-field measurements have been
carried out in a free flight analy-sis laboratory using the PIV (Particle
Image Velocimetry) technique. Instantaneous velocity fields were measured in
successive planes crosswise to the flight path of a free flying air-liner
model, so that the evolution, interaction and decay of vortices originating
from the wing, flap and slat tips can be described. The optical measurement
sys-tem consisting of two aerosol generators for seeding, a pulsed laser
system for illumination, two CCD cameras for image recording and trigger
electronics for synchroni-sa-tion had to be adapted to the particular test
conditions of a catapult facil-ity, where a long-lasting, unsteady "single
event" initiated by the model's launch is subject of the experi-mental
study. The paper focuses on the tech-nical aspects of interfacing a catapult
facility and a digital PIV system.
- 2
-
L. Dieterle, R. Stuff, H. Vollmers, and P. Coton.
Wake vortex studies in the ONERA catapult facility and in the DNW
wind tunnels.
In (ODAS) ONERA DLR Aerospace Symposium, Paris (France), 21 -
24 June 1999.
As part of the European research programmes WAVENC and EUROWAKE the
wake vortex system behind (i) a free flying Airbus model and (ii) the DLR's
ALVAST half model has been investigated experimentally in the ONERA catapult
facility and in the DNW-LLF (large low-speed wind tunnel facility)
respectively. Digital particle image velocimetry (PIV) was used for both
testing campaigns. The two facili-ties offer different experimental access to
the physi-cal phenomena of vortex formation, evolution and decay with one and
the same measurement technique and require a particular adaptation of this
measurement technique to the different test conditions each. In both cases
tracer particles added to the flow were illuminate d by a pulsed laser light
sheet in a plane crosswise to the flight path of the model or to the main
flow direction respectively. Two CCD cameras recorded neighbouring areas of
the flow field simultaneously. The PIV system operation was fitted to the
facilities' requirements. Image data were evaluated utilising the
cross-correlation method. The resulting velocity and vorticity fields give
quantitative information on the structure and motion of the trailing vortices
in different frames of reference: a ground-based one for the catapult and a
model's one for the wind tunnel.
- 3
-
L. Chaudron.
Se comprendre pour travailler ensemble.
In Enac, editor, Colloque "Les facteurs humains en
aéronautique", pages 57-66, Toulouse (France), 7 - 8 juin 1999. ENAC.
Deux contrôleurs coopèrent pour le suivi d'un ensemble d'avions, deux
mathématiciens discutent de la démonstration d'un théorème, deux critiques
confrontent leur point de vue afin de rédiger un article commun, deux
philosophes débattent d'une thèse, deux pilotes collaborent pour l'exécution
de leur mission... dans toutes ces différentes situations, plusieurs agents,
rationnels, dialoguent avec comme but commun la réalisation, immédiate ou
différée, d'un «opus» (le résultat de leur co-opération) qui dépend
fondamentalement de la mise en commun de toutes leurs informations. Cette
question de la synthèse d'information et de connaissance, apparaît comme une
composante fondamentale des organisations au sein desquelles doivent
travailler ensemble aussi bien des êtres humains que des systèmes
artificiels. Le but de cet article est d'examiner quelques problèmes
essentiels qui se présentent naturellement dès que la question du partage de
connaissances entre différents agents - on appellera "agent" aussi bien
un être humain qu'un système artificiel doué des facultés de perception,
raisonnement et action - se pose.
- 4
-
N. Bataille, M. Lemaître, and G. Verfaillie.
Efficiency and fairness when sharing the use of a satellite.
In 5th International Symposium on Artificial Intelligence,
Robotics and Automation in Space (iSAIRAS-99), pages 465-470, Noordwijk
(NL), 1 - 3 June 1999. ESTEC.
Resources co-funded by several agents must be exploited in such a way
that three kinds of constraints are met: (1) physical problem (hard)
constraints; (2) efficiency constraints, aiming at maximizing the
satisfaction of each agent; (3) a fairness constraint, which is ideally
satisfied when each agent receives an amount of the resource exactly
proportional to its financial contribution. This paper investigates a
decision problem for which the common property resource is an earth
observation satellite. The problem is to decide on the daily selection of a
subset of pictures, among a set of candidate pictures which could be taken
the next day considering the satellite trajectory. This subset must satisfy
the three kinds of constraints stated above. Although fair division problems
have received considerable attention for a long time, especially from
microeconomists, this specific problem does not fall entirely within a
classical approach. This is because the candidate pictures may be
incompatible, and because a picture is only of value to the agent requesting
it. As in the general case, efficiency and fairness constraints are
antagonistic. We propose three ways for solving this share problem. The first
one gives priority to fairness, the second one to efficiency, and the third
one computes a set of compromises.
- 5
-
L. Lobjois and M. Lemaître.
LNS/VSP: une méthode de recherche locale pour la résolution de
VCSP en contexte interruptible.
In Journées Nationales sur la Résolution Pratique de Problèmes
NP-Complets (JNPC99), pages 163-170, Lyon (France), 2 - 4 juin 1999.
Dans cet article, nous proposons une nouvelle méthode hybride entre
méthodes complètes et méthodes incomplètes pour résoudre, dans un contexte
interruptible, les problèmes d'optimisation combinatoire exprimés dans le
cadre Valued Constraint Satisfaction Problem (VCSP). Le principe de la
méthode proposée est d'utiliser, au sein d'une recherche locale utilisant des
voisinages étendus, des mécanismes de propagation de contraintes issus des
méthodes complètes. Nous présentons des résultats expérimentaux montrant
l'intérêt d'une telle approche dans un contexte interruptible (anytime),
c'est à dire lorsque la résolution peut être stoppé à tout moment.
- 6
-
D. Farcy.
Manoeuvrability of submarines.
In paper N° 12, London UK, 14 - 16 June 1999. Royal Institution
of Naval Architects.
This paper deals with a numerical method initially developed for
predicting high-angle-of-attack aircraft manœuvres and applied to
submarine behaviour analysis. Experimental tests are required to determine
the hydrodynamic forces and to build a mathematical modelling. The methods
undertaken at ONERA in this field are presented. The bifurcation theory is
devoted to the analysis of the steady-state response and stability of
non-linear systems. First a few aspects of this theory are described. By use
of bifurcation theory a comprehensive non linear study of the equilibria and
stability of submarines is realised. The behaviour even at the boundaries of
the operational envelope is analysed, and losses of stability or other
dangerous behaviours are determined in function of the controls or
parameters. Variation of buoyancy appears as the most sensitive factor.
Results are completed and illustrated by numerical simulations of
manœuvres.
- 7
-
G. Ferreres and C. Döll.
Robust pole placement for an LFT model.
In ACC 99, pages 2233-2237, San Diego, California, USA, 2 - 4
June 1999. AACC, IEEE.
A method is proposed for synthesizing a controller, whose order is a
priori fixed, and which ensures a robust pole placement inside a given region
of the complex plane, despite parametric uncertainties in the plant model.
The key idea is to use a dyadic parameterization of the controller, so as to
transform the non-convex optimization problem into a series of Linear
Programming problems. The method is successfully applied to the space shuttle
example presented in the User's Guide of the Matlab mu-analysis and synthesis
Toolbox.
- 8
-
H. D. Tuan, E. Ono, P. Apkarian, and S. Hosoe.
Nonlinear hinfini control for an integrated suspension system via
parameterized linear matrix inequality characterizations.
In 1999 American Control Conference (ACC), San Diego (USA), 2 -
4 June 1999. IEEE.
The automotive hydro-pneumatic integrated suspension model is
nonlinear with large dimensions. As a consequence, the nonlinear Hinfinity
control methodology based on the traditional Hamilton-Jacoby-Isaacs equation
is impractical in this application. An alternative so-called Parameterized
Linear Matrix Inequality (PLMI) approach is proposed for solving this hard
nonlinear Hinfinity control problem. The validity of the proposed approach is
confirmed not only by detailed and realistic simulations but also by
extensive experiments. Specifically, the proposed nonlinear control method
outperforms the more classical feedback linearization control technique.
- 9
-
P. Apkarian and H. D. Tuan.
LMI-constrained concave programs in robust control.
In ACC 1999, San Diego (USA), 2 - 4 June 1999. IEEE.
We show in the present paper that many open and challenging problems
in control theory belong to the class of concave minimization programs. More
precisely, these problems can be recast as the minimization of a concave
objective function over convex LMI (Linear Matrix Inequality) constraints. In
this setting, these problems can then be efficiently handled using local
and/or global optimization techniques.
- 10
-
P. Apkarian, H. D. Tuan, and Y. Nakashim.
A new lagrangian dual global optimization algorithm for solving
bilinear matrix inequalities.
In IEEE, editor, ACC 1999, San Diego (USA), 2 - 4 June 1999.
IEEE.
A new global optimization algorithm for solving Bilinear Matrix
Inequalities (BMI) problems is developed. It is based on a dual Lagrange
formulation for computing lower bounds that are used in a branching procedure
to eliminate partition sets in the space of nonconvex variables. The
advantage of the proposed method is twofold. First, lower bound computations
reduce to solving easily tractable Linear Matrix Inequality (LMI) problems.
Secondly, the lower bounding procedure guarantees global convergence of the
algorithm when combined with an exhaustive partitioning of the space of
nonconvex variables. A rigorous proof of this fact is provided. Another
important feature is that the branching phase takes place in the space of
nonconvex variables only, hence limiting the overall cost of the algorithm.
Also, an important point in the method is that separated LMI constraints are
encapsulated into an augmented BMI for improving the lower bound
computations. Applications of the algorithm to robust structure/controller
design are considered.
- 11
-
G. Verfaillie, E. Bensana, C. Michelon-Edery, and N. Bataille.
Dealing with uncertainty when managing an observation satellite.
In 5th International Symposium on Artificial Intelligence,
Robotics and Automation in Space - iSAIRAS, pages 205-207, Noordwijk (NL),
1 - 3 June 1999. ESTEC.
The possible presence of clouds is the main origin of uncertainty
when managing earth optical observation satellites. Forgetting it can lead to
poor results in terms of really achieved photographs. In this paper, we show
how a mathematical approach, drawn from the Markov Decision Process
framework, allows us to define a rational way of taking in account this
uncertainty in the daily optimization process. Keywords : planning,
uncertainty, markov decision process.
- 12
-
G. Verfaillie and L. Lobjois.
Problèmes incohérents: expliquer l'incohérence, restaurer la
cohérence.
In Journées Nationales sur la Résolution Pratique de Problèmes
NP-Complets (JNPC99), pages 111-120, Lyon (France), 2 - 4 juin 1999.
Si vous utilisez un outil de programmation par contraintes et que,
pour une raison ou une autre, le problème que vous lui soumettez est
incohérent, la seule réponse qu'en général vous obtenez est incohérence, ce
qui est somme toute normal, mais a parfois le don de vous plonger dans un
abîme de perplexité. Quelles sont les raisons de cette incohérence? Que faire
pour restaurer la cohérence ? À ces légitimes questions, votre outil préféré
ne vous apporte aucune réponse. Dans ce papier, nous proposons une
formalisation de ces deux requêtes, dans le cadre des Problèmes de
Satisfaction de Contraintes (CSP). Nous les situons l'une par rapport à
l'autre et nous évaluons leur complexité. C'est en particulier l'occasion de
montrer que la plus complexe n'est pas celle qu'on croit. Enfin, nous
montrons comment des techniques connues ou simples à définir et à mettre en
œuvre peuvent apporter des réponses au moins partielles à ces deux
requêtes.
- 13
-
S. Prudhomme, C. Blondeau, M. Humbert, and A. Bucharles.
An unsteady aerodynamics identification procedure for flutter
prediction.
In International Forum an Aeroelasticity & Structural Dynamics
1999, Willamsburg (USA), 22 - 25 June 1999. NASA.
This article introduces an identification procedure from flight data
in the perspectives of quasi real time flutter prediction modes upgrading.
Parameters are introducted to model uncertainties on the unsteady
aerodynamics. While tests are performed for increasing velocity,
identification algorithms use flight data to update the parameters and make
flutter prediction more accurate for investigation towards higher velocity.
Performance of identification is evaluated in terms of accuracy for unsteady
aerodynamics estimation, and of flutter prediction capability with and
without turbulence.
|
