Learning to Solve Stochastic Multi-Agent Path Finding

Guillaume Dalle (1) , Axel Parmentier (1)

1 CERMICS - Centre d'Enseignement et de Recherche en Mathématiques et Calcul Scientifique

In large transportation networks, real-time traffic management is essential to minimize disruptions and maximize punctuality. This is especially true for railway systems, where delays can easily propagate from one train to the next due to infrastructure constraints. We propose novel algorithms to tackle this problem, using the Flatland challenge as a testing ground.

Mots clés: online stochastic optimization; vehicle rescheduling; multi agent path finding; machine learning for operations research

Domaines: Recherche opérationnelle [math.OC]; Combinatoire [math.CO]

Guillaume Dalle, Axel Parmentier. Learning to Solve Stochastic Multi-Agent Path Finding. 23ème congrès annuel de la Société Française de Recherche Opérationnelle et d'Aide à la Décision, INSA Lyon, Feb 2022, Villeurbanne - Lyon, France. ⟨hal-03595315⟩