Decomposition-Coordination for dynamic stochastic optimization problems: Theory and Applications of the DADP algorithm
31 Outubro 2013 ·
- Quem: Vincent Leclere
- Onde: FGV - Praia de Botafogo, 190, room 317
- Quando: 31 de Outubro de 2013 às 16:00h
Large scale stochastic optimization problems are too difficult to be solved directly numerically. One way of dealing with this type of problems is to decompose the main problem in smaller one and coordinate the solutions of each smaller problems. In a deterministic setting spatial decomposition methods are well known but quite useless if directly used in a stochastic setting. We present the Dual Approximate Dynamic Programming algorithm which solve an approximation of the original problem by decomposition. We will present the algorithm, some thoughts and results about its convergence and numerical application and results over an energy management problem.
Biography
Vincent Leclere has specialization in Applied Mathematics (École Polytechnique (2010)), Master in Game Theory and Optimization, Master in Finacial Mathematics (Université Paris) and currently PhD student in Stochastic Optimization of discrete-time dynamic system.
Important
- Not required a confirmation.
- FGV not allow the entrance of people wearing shorts and/or slippers.