Using a Monte-Carlo Approach to Bus Regulation

Tristan Cazenave (1) , Flavien Balbo (1) , Suzanne Pinson (1)

1 LAMSADE - Laboratoire d'analyse et modélisation de systèmes pour l'aide à la décision

In this paper, our objective is to minimize passengers waiting times at the bus stops by making buses wait at a stop. We compare a simple rule based approach to a Monte-Carlo method to reach this objective. When allocated enough time, the Monte-Carlo method gives better results. If the passengers arrivals and the bus travel times are known, the best algorithm is nested Monte-Carlo search with memorization which clearly outperforms nested Monte-Carlo search without memorization as well as Monte-Carlo and rule based regulation.

Mots clés: Bus regulation; Monte Carlo methods; Road vehicles; traffic engineering computing; Intelligent transportation Systems; Traffic control; Agent systems; Rule-based Systems

Domaines: Informatique [cs]

Tristan Cazenave, Flavien Balbo, Suzanne Pinson. Using a Monte-Carlo Approach to Bus Regulation. 12th IEEE Intenrational Conference On Intelligent Transportation Systems (ITSC'09), Oct 2009, St-Louis, United States. ⟨10.1109/itsc15659.2009⟩. ⟨hal-03695449⟩