Project-team OPALE INRIA Sophia-Antipolis Mditerrane and - - PowerPoint PPT Presentation

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Project-team OPALE

Optimization and control, numerical algorithms and integration

• f complex multidisciplinary systems governed by PDEs

Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes NumD Evaluation Seminar, Paris, March 17-19, 2009

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Overview

1

Scientific Themes

2

Personnel

3

Main achievements Objective 1: Numerical methods for simulation and shape

• ptimization

Objective 2: Mathematical analysis of geometrical optimization Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

4

Objectives for the next four years Objective 1: Numerical methods for simulation and shape

• ptimization

Objective 2: Mathematical analysis of geometrical optimization Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

5

Positioning w.r.t. INRIA’s priorities

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Overview

1

Scientific Themes

2

Personnel

3

Main achievements Objective 1: Numerical methods for simulation and shape

• ptimization

Objective 2: Mathematical analysis of geometrical optimization Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

4

Objectives for the next four years Objective 1: Numerical methods for simulation and shape

• ptimization

Objective 2: Mathematical analysis of geometrical optimization Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

5

Positioning w.r.t. INRIA’s priorities

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Scientific Themes, Ambitions (1)

Project-Team created in 2002 by J.-A. DESIDERI (INRIA SOPHIA ANTIPOLIS)

• A. HABBAL (UNSA)
• T. NGUYEN (INRIA GRENOBLE)

J.-P. ZOLESIO (CNRS SOPHIA ANTIPOLIS)

Numerical Optimization of PDE systems

MOSTLY (=NOT EXCLUSIVELY) DESIDERI+DUVIGNEAU

• Develop, analyze and experiment numerical methods for

PDE-constrained optimization, and in particular optimum-shape design in complex scientific applications

• Multi-disciplinary context: Computational Fluid Dynamics (CFD),

Electromagnetics (CEM), Structural Mechanics (CSM), Biology, and couplings

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Scientific Themes, Ambitions (2)

Mathematical analysis of geometrical optimization

MOSTLY ZOLESIO AND CO.

• Main themes:
• Large evolution of geometry
• Transverse field analysis
• Shape stabilization for wave equn. in moving domain
• Tube analysis and geodesics in relation to Topological Evolution
• Main applications:
• Electromagnetics
• Fluid-structure interaction
• Foundations:
• Shapes and Geometries, Delfour-Zolésio, SIAM
• Zolésio-Sokolowski
• Cagnol-Zolésio

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Scientific Themes, Ambitions (3)

Computing Environments

MOSTLY NGUYEN AND CO.

• Scientific objectives
• High performance computing
• Problem solving environments
• Coupling distributed, parallel, heterogeneous codes
• Uptake in large-scale multidiscipline applications
• Targeted technologies
• Collaborative integration platforms
• Distributed high-performance problem solving

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Scientific Themes, Ambitions (4)

Application of shape and topology design to biology and medecine

MOSTLY HABBAL AND CO.

• Main objective
• Model biological phenomena with mathematical tools in narrow relation with shape

and topological optimization theory

• Domains under study
• Anti-angiogenesis
• Wound healing

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Overview

1

Scientific Themes

2

Personnel

3

Main achievements Objective 1: Numerical methods for simulation and shape

• ptimization

Objective 2: Mathematical analysis of geometrical optimization Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

4

Objectives for the next four years Objective 1: Numerical methods for simulation and shape

• ptimization

Objective 2: Mathematical analysis of geometrical optimization Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

5

Positioning w.r.t. INRIA’s priorities

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Scientific and Technical Staff (March 2009) (1)

Permanent researchers

• Jean-Antoine Désidéri (DR, INRIA Sophia Antipolis), Head
• Régis Duvigneau (CR, INRIA Sophia Antipolis)
• Abderrahmane Habbal (Asst. Prof., University of Nice Sophia Antipolis)
• Toan Nguyen (DR, INRIA Grenoble)
• Jean-Paul Zolésio (DR CNRS, INRIA Sophia Antipolis)

Staff on secondment

• Paola Goatin (Asst. Prof., University of Toulon) - part time

Post-doctoral fellows

• Abderrahamane Benzaoui (French ANR Project OMD)
• Xavier Hachair (Thalès grant)
• Luigi Manca (INRIA grant)
• Mohammed Ziani (European Project EXCITING)

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Scientific and Technical Staff (March 2009) (2)

Doctoral student

• Benoît Chaigne (France Telecom/PACA Region grant)

Technical staff

• Nikolas Mauny (INRIA grant)

Collaborators

• Prof. John Cagnol (Pôle Universitaire Léonard de Vinci, Paris)
• Asst. Prof. Praveen Chandrashekarappa (Tata Institute, Bangalore, India)
• Prof. Michel Delfour (Centre de Recherche en Mathématiques, Montreal, Canada)
• Prof. Raja Dziri (University of Tunis, Tunisia)
• Prof. Jacques Périaux (University of Jyväskyla, Finland, and Universitat Politecnica de

Catalunya, Barcelona, Spain)

• Prof. Jan Sokolowski (University Henri Poincaré Nancy I)

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Overview

1

Scientific Themes

2

Personnel

3

Main achievements Objective 1: Numerical methods for simulation and shape

• ptimization

Objective 2: Mathematical analysis of geometrical optimization Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

4

Objectives for the next four years Objective 1: Numerical methods for simulation and shape

• ptimization

Objective 2: Mathematical analysis of geometrical optimization Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

5

Positioning w.r.t. INRIA’s priorities

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Objective 1 (2005-2009)

Numerical methods for simulation and shape

• ptimization of multidisciplinary PDE systems
• Multilevel shape optimization algorithms
• Hybrid optimizers
• Coordination of physical models
• Neural networks and Kriging methods
• Robust design

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Multilevel shape optimization

• Basic validation of concept1
• Analysis of algebraic model2
• Size experiments in compressible aerodynamics3,4
• Parameterization self-adaption procedures5
• Multilevel shape optimization of antennas6
• Stochastic/deterministic Hybridization7
• Software: FAMOSA platform + Scilab toolbox
• Participation in two European short courses on
• ptimization (ERCOFTAC, Von Karman Institute)
• Invited conference at the German Aerospace Lab

(DLR Braunschweig)

• On-going: extension to algebraic hierarchical basis

1

• J. Computational Physics, 2007

2 Advances in Numerical Mathematics, 2006 3

• B. Abou El Majd’s Doctoral Thesis, 2007

4 European J. of Computational Mechanics, 2008 5 European Series in Applied and Industrial Mathematics, 2007 6

• B. Chaigne’s Doctoral Thesis, 2009

7 Optimisation Multidisplinaire en Mécanique, Hermès, 2009

stiff + multimodal problem

= ⇒

multilevel + hybrid + parallel algorithm

fine medium coarse

PSO simplex simplex PSO PSO simplex simplex

Free-Form Deformation

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Coordination of physical models

Game strategies with sensible split of territory

• Proof of concept in compressible aerodynamics1
• Theoretical split for a preponderant discipline2
• Aero-structural shape optimization of an aircraft

wing3

• Nash and Stackelberg games
• Coupling of High and Low-fidelity models
• Theoretical extension to the general case4

(on-going)

• Invitation to Workshop on PDE-Constrained

Optimization, Oberwolfach

1

• J. of Optimization Theory and Applications (JOTA), 2007

2 INRIA Report 6108, 2007 (english), or: Optimisation Multidisplinaire en Mécanique, Hermès, 2009 (french) 3

• B. Abou El Majd’s Doctoral Thesis, 2007

4 Mathematishes Forschungsinstitut Oberwolfach Report, 2009

Aerodynamics optimized alone Aero-structural Nash equilibrium

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Neural networks and Kriging

• Prize for best communication

(Meta-modelling for multi-level

• ptimization and robust design1)
• Acceleration of PSO using inexact

pre-evaluations2

• Gaussian-process-based optimization3
• Gradient+Hessian estimates for Nash

games4

• Inexpensive models for Monte-Carlo

simulation5

• High and Low-Fidelity Model Coupling for

MDO6

1 Applied Aerodynamics Colloquium (AAAF), 2007 2 Computer Methods in Applied Mechanics and Engineering, 2009 3 Applied Aerodynamics Colloquium (AAAF), 2009 4 Optimisation Multidisplinaire en Mécanique, Hermès, 2009 5 10th AIAA Non-Deterministic Approaches Conference, 2008 6

• B. Abou El Majd’s Doctoral Thesis, 2007

Gaussian process model for drag minimization

• f oscillatory-rotating cylinder

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Robust design

Uncertainty estimation

• Second derivatives of cost-functional w.r.t.

uncertain variables (Automatic Differentiation TAPENADE)1

• Metamodel-based Monte-Carlo2
• Uncertainty estimation over flowfields using the

sensitivity-equation method3

Optimization under uncertainty4

• Multipoint
• Worst-case
• Interval analysis
• Statistical (probabilistic)

1 10th AIAA Non-Deterministic Approaches Conference, 2008 2 Optimisation Multidisplinaire en Mécanique, Hermès, 2009 3 International J. of Computational Fluid Dynamics, 2006 4 Applied Aerodynamics Colloquium (AAAF), 2009

Wing shape optimized at M=0.83 and recomputed at M=0.85 Wing shape from robust optimization and recomputed at M=0.85

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Objective 2 (2005-2009)

Mathematical analysis of geometrical optimization

• Dynamical domain control
• In coupled fluid-structure devices1
• Morphic metric in Shape Space2
• Strong solution to the Hamilton Jacobi equation for topological change3
• Habilitation of J. Ferchichi (now Associate Prof. at Monastir)4
• Shape stabilization of wave equation
• Stabilization in intrinsic shell equations: on-going with I. Lasiecka (UVA)
• Habilitation of J. Cagnol (now prof. at PULV)5

1 Monograph: Moving Shape Analysis (Chapman & Hall), 2006 2 23rd IFIP TC 7, Cracow (P), Springer Verlag, 2009 (to appear) 3 International Series of Numerical Mathematics, 2007 4 Free boundary control in visco-elastic fluid (Monastir), 2007 5 Control and stabilization of elastic shells, PULV, 2006

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Objective 2 (2005-2009)

e-lab OpRaTel

• Partners : Opale-Thales-France Telecom
• Theme : Optimization of antennas
• Setting : control in three-dimensional electromagnetics

for radiating solution in near field as well as far field. For both the time-dependent solution to the Maxwell equations and for the harmonic regime including the integral representation of solution via the variational principle associated with the reciprocity

• property. Shape gradient in a full mathematical setting including the bridge between the

two theories (harmonic and Maxwell), and the derivative with respect to the geometry of the Colton Kreiss isomorphism on the boundary for the harmonic approach.

• Shape reconstruction of 3D dialectric obstables by level-sets1
• Optimization of array antenna devices2

1 P . Dubois, doctoral thesis, 2005 2

• L. Blanchard, doctoral thesis, 2007

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Objective 3 (2005-2009)

Computing environments

• Design, deployment and experimentation of distributed

high-performance computing environments

• Simplification of the access to distributed computing resources
• Virtualization techniques for the design and deployment of

distributed workflow systems

• Contribution to scientific networking in National and

International projects

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Distributed computing

• High-level user interface1
• Middleware support: Globus, UNICORE2
• Interface with standards: CORBA, J2EE,

WSRF

• Implemented with CAST software3
• Distributed multi-cluster computing4
• Cooperation with CNES in HEAVEN

consortium Fluid-structure interaction example Fluid-structure interaction workflow

1 Third International Conference on Computing, Communications and Control Technologies (CCCT’05), Austin, Texas (USA), 2005 2 Second Grid@Asia Workshop Shanghai (China), 2006 3 Tenth CSCW in Design Conference, Nanjing (China), 2006 4 ECCOMAS 2nd Intl Conference on Multiphysics Coupled Problems, Ibiza (Spain), 2007

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

High-performance computing

• Virtual collaborative platform
• Using distributed workflow

systems1,2

• Current transfer to YAWL

workflow system

• Aiming at multidiscipline

applications3

• Cooperation in AEROCHINA 1

and 2 networks Wing optimization workflow

1 International Conference on CSCW in Design (CSCWD 2008), Xi’an (China), 2008 2 Fifth European Congress on Computational Methods in Applied Sciences and Engineering (ECCOMAS 2008), Venice (I), 2008 3 Fifth Intl. Conference on Cooperative Design, Visualization and Engineering CDVE 2008, Mallorca (Spain), 2008

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Objective 4 (2005-2009)

Application of shape and topology design to biology and medecine

• Anti-angiogenesis for solid tumors
• Wound healing

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Tumoral anti-angiogenesis modeled as a Nash game

Activators strategy Host vessel=upper side Inhibitors strategy Tumour=inner circle Final Network at Nash equilibrium

• Model: Elastic-porous media models coupled via the interaction of

activator/inhibitor densities (topology Nash game)

• Results: theoretical existence of a NE, computational issues, multiple

channels structures arise, not arterial trees1,2

1 Journal of Structural Multidisciplinary Optimization, 2005 2 Revue Africaine de la Recherche en Informatique et Mathématiques Appliquées, 2008

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Embryonic wound dynamics I. Dorsal closure with zipping

Scheme of tissues and forces implicated in dorsal closure in Drosophila embryos. Dorsal Closure images at successive times with simulated contours (in red).

• Model: quasi-static linear elasticity, with a geometric flow and zipping
• perator, level sets drive the wound evolution
• Results: model validated, parameter identification, good capturing of

the purse string mechanism1

• Collaborators: L. Almeida, CNRS ; S. Noselli, F. Serman, Biochimie
• Univ. Nice & INSERM ; P

. Bagnerini, Univ. Genova, IT.

• 1. Singularities in nonlinear evolution phenomena and applications, CRM Series, S. Normale S., Pisa, 2008.

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Embryonic wound dynamics II. Closure after laser ablation

Wound images at successive times with extracted (blue) and simulated (red) contours. Time increases from left to right and top to bottom.

• Model: quasi-static linear elasticity, with normal and mean-curvature

flows, level sets drive the wound evolution

• Results: model well validated1, currently used to classify Drosophila

mutants

• Collaborators: L. Almeida, CNRS ; S. Noselli, F. Serman, Biochimie
• Univ. Nice & INSERM ; P

. Bagnerini, Univ. Genova, IT.

• 1. Singularities in nonlinear evolution phenomena and applications, CRM Series, S. Normale S., Pisa, 2008.

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Wound healing III. MDCK cell monolayers

Different time steps comparing the closure of an MDCK monolayer (up) to simulation results from the Fisher-KPP equation (bottom).

• Model: 2D Fisher-KPP (parabolic) PDE, image processing
• Results: computational evidence shows that "wounds close at

constant speed",

• Trends: integrating visco-elasticity and cell-matrix interaction

(hyperbolic PDE) in Oster-Murray-type model

• Collaborators: G. Malandain, INRIA ; H. Barelli, IPMC CNRS &

INSERM.

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Overview

1

Scientific Themes

2

Personnel

3

Main achievements Objective 1: Numerical methods for simulation and shape

• ptimization

Objective 2: Mathematical analysis of geometrical optimization Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

4

Objectives for the next four years Objective 1: Numerical methods for simulation and shape

• ptimization

Objective 2: Mathematical analysis of geometrical optimization Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

5

Positioning w.r.t. INRIA’s priorities

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Objective 1

Numerical methods for simulation and shape

• ptimization of multidisciplinary PDE systems
• Multilevel algorithms in abstract hierarchical basis
• Iso-geometric approaches
• General algorithms for multi-disciplinary
• ptimization
• Uncertainties (estimation and prapagation) and

robust design

• Self-adaptive algorithms: integration of

metamodel-based strategies

• New applications and reinforced collaborations:
• finite-volumes for pedestrian traffic
• concurrent engineering for racing boat

design

• M.I.T., University of Jyväskyla
• ONERA
• Arcelor Mittal
• Renault

NURBS-BASED GRID AND SOLUTION

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Objective 2

Mathematical analysis of geometrical optimization

• Dynamic domain control in coupled fluid-structure devices:

Derive the optimal shape evolution of a boundary for minimal drag/curl in the non cylindrical Navier-Stokes.

• Hidden regularity (sharp) in electromagnetics.

The challenge is to understand the bridge between Maxwell evolution system and boundary integral equation for harmonic radiating solutions in the presence of boundary singularity on the non metallic obstacle

• e-lab OpRaTel

The development of the new shape “morphic” metric with geosdesic construction is under exploitation with Thales for radar detection and the post doc X. Hachair is supported . Another promising development is done with Luigi Manca using optimal control approach for stable numerical methods for harmonic regime control.

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Objective 3

Computing environments

• Dynamic resilient distributed computing environments
• Configurable distributed workflow systems
• Interface of high-performance infrastructures to distributed

workflow systems

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Objective 4

Cell and Tissue Dynamics

• Embryonic wound dynamics: visco-elastodynamics,

long-range zipping kernels

• MDCK monolayer model: relate vesicle trafficking

(pharmacokinetics) to wound-closure dynamics

• Games to model mean-field behaviour of multicellular

populations

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Overview

1

Scientific Themes

2

Personnel

3

Main achievements Objective 1: Numerical methods for simulation and shape

• ptimization

Objective 2: Mathematical analysis of geometrical optimization Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

4

Objectives for the next four years Objective 1: Numerical methods for simulation and shape

• ptimization

Objective 2: Mathematical analysis of geometrical optimization Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

5

Positioning w.r.t. INRIA’s priorities

Project-team OPALE Optimization and control, numerical algorithms and integration of complex multidisciplinary systems governed by PDEs Jean-Antoine DESIDERI INRIA Sophia-Antipolis Méditerranée and Rhône-Alpes Scientific Themes Personnel Main achievements

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Application of shape and topology design to biology and medecine

Objectives for the next four years

Objective 1: Numerical methods for simulation and shape optimization Objective 2: Mathematical analysis

• f geometrical optimization

Objective 3: Computing environments Objective 4: Cell and Tissue Dynamics

Positioning w.r.t. INRIA’s priorities

Positioning w.r.t. INRIA’s priorities

W.r.t. INRIA’s seven strategic priorities (2008-2012)

• Modeling, Simulation and Optimization of Complex Dynamic

Systems

• Information, Computation and Communication Everywhere
• Computational Engineering
• Computational Sciences