Séminaire OM/PMNL


Séminaire Optimisation Mathématique - Programmation Mathématique Non Linéaire
Jeudi, 10 Juin, 2021

Dans le cadre du GdR-RO, l’axe OM/PMNL (Optimisation Mathématique - Programmation Mathématique Non Linéaire) propose le jeudi 10 juin prochain de 14h à 16h deux exposés en Visio-conférence.

Nous aurons ainsi le plaisir d’accueillir : 

  • Miguel F. Anjos,  Professeur à l’Université d’Edinburgh -  14h
    • Title: AC Optimal Power Flow and Optimal Reactive Power Dispatch: Successes and Challenges for the use of Semidefinite Optimization
    • Abstract: The classical alternating current optimal power flow problem is nonconvex and generally hard to solve to global optimality. There has been a lot of interest in the past decade for the application of semidefinite optimization to this problem, and more recently to the related problem of optimal reactive power dispatch. This presentation will survey the main results on this application and highlight some of the challenges still outstanding. 
  • Ignacio Araya,  Professeur à l'Université du Chili - 15h
    • Title:  Interval-based methods for constrained global optimization
    • Abstract: Interval-based methods can be used to solve rigorously constrained global optimization problems. Two main features distinguish interval-based methods from other methods: (1) they can take into account uncertainties in the parameter values and computation errors over the floating-point numbers and (2) they are able to treat the complete search space, thus offering proofs of infeasibility and/or certification of solutions. In this presentation, I will explain the basic principles of intervals and interval-based methods for solving constrained global optimization problems. Then,  I will talk about the different components of interval branch and bound algorithms, focusing on two particular approaches: an inner region-based method for finding feasible solutions and a Lagragian-based method for selecting the variable to split. Finally, I will show how we can adapt the solver to biobjective optimization problems.

Les animateurs de l’axe OM/PMNL

Amélie Lambert, Sonia Cafieri, Claudia D’Ambrosio, Frédéric Roupin, Gilles Trombettoni, Frédéric Messine.