Integer programming for weakly coupled stochastic dynamic programs with partial information - École des Ponts ParisTech Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2020

Integer programming for weakly coupled stochastic dynamic programs with partial information

Résumé

This paper introduces algorithms for problems where a decision maker has to control a system composed of several components and has access to only partial information on the state of each component. Such problems are difficult because of the partial observations, and because of the curse of dimensionality that appears when the number of components increases. Partially observable Markov decision processes (POMDPs) have been introduced to deal with the first challenge, while weakly coupled stochastic dynamic programs address the second. Drawing from these two branches of the literature, we introduce the notion of weakly coupled POMDPs. The objective is to find a policy maximizing the total expected reward over a finite horizon. Our algorithms rely on two ingredients. The first, which can be used independently, is a mixed integer linear formulation for generic POMDPs that computes an optimal memoryless policy. The formulation is strengthened with valid cuts based on a probabilistic interpretation of the dependence between random variables, and its linear relaxation provide a practically tight upper bound on the value of an optimal history-dependent policies. The second is a collection of mathematical programming formulations and algorithms which provide tractable policies and upper bounds for weakly coupled POMDPs. Lagrangian relaxations, fluid approximations, and almost sure constraints relaxations enable to break the curse of dimensionality. We test our generic POMDPs formulations on benchmark instance forms the literature, and our weakly coupled POMDP algorithms on a maintenance problem. Numerical experiments show the efficiency of our approach.

Dates et versions

hal-03040397 , version 1 (04-12-2020)

Identifiants

Citer

Victor Cohen, Axel Parmentier. Integer programming for weakly coupled stochastic dynamic programs with partial information. 2020. ⟨hal-03040397⟩
26 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More