3e cycle romand de Recherche Opérationnelle
March 4-8, 2001, Hotel de l'Europe, Zinal (VS), Switzerland
Spring seminar / Séminaire de printemps
Information and registration : http://rosowww.epfl.ch/3emecycle or alain.prodon@epfl.ch
Registration's deadline : January 31, 2001
Invited lecturers :
Aharon Ben-Tal, Technion - Israel Institute of Technology (Israel) and TU - Delft (NL)
Modern Convex Optimization in Engineering DesignMichael C. Ferris, University of Wisconsin (USA)
Formulation, Modeling and Solution : a Nonlinear Perspective
Talk Abstracts :
Modern Convex Optimization
in Engineering Design
Aharon Ben-Tal
Truss Topology Design(TTD)
ˇ Introduction to trusses ˇ Sizing, topology and geometry design ˇ Simple and multiple load TTD ˇ Mathematical modeling of the TTD problem ˇ Equivalent convex programming formulations ˇ Examples of optimal trusses ˇ Stability of optimal trusses
Antenna and Filter Design
ˇ Introduction to filters and antennae ˇ Design problems associated with filters and antennae: ˇ Synthesis of array of antennae ˇ Low pass filter ˇ Stability issues
Robust Optimization(RO)
ˇ The RO-methodology ˇ The need for RO: unstable solutions of engineering design problems ˇ The Linear Programming NETLIB case study ˇ Basic theory of robust LP, Conic Quadratic and Semi Definite programming ˇ The engineering design problems revisited; robust TTD and Robust antenna design ˇ The NETLIB case study revisited
Formulation, Modeling
and Solution : a Nonlinear Perspective
Michael C. Ferris
ˇ Complementarity: an overview ˇ Complementarity modeling ˇ Large scale complementarity solvers ˇ Visualisation - GAMS/Matlab ˇ The Gamma Knife ˇ Simulation optimization ˇ Support vector machine applications ˇ Metacomputing in optimization
There is a general focus that they involve "nonlinear problems and modeling".
The first three relate to complementarity, describing first the applications and the formulation, then how to model these, then how to solve them.
After that, I would like to present some tools on visualisation of results from modeling languages. The remainder would look at some nonlinear applications, one to brain tumor treatment, another in simulation optimization, and finally to SVM's. I would finish with a view of computing environments for the future.