Reinforcement learning: Introduction to theory and potential for transport applications

Abstract

The aim of this paper is to develop insight into the potential of reinforcement learning (RL) agents and distributed reinforcement learning agents in the domain of transportation and traffic engineering and specifically in Intelligent Transport Systems (ITS). This paper provides a crystallized, comprehensive overview of the concept of RL and presents related successful applications in the field of traffic control and transportation engineering. It is divided into two parts: the first part provides a thorough overview of RL and its related methods and the second part reviews most recent applications of RL algorithms to the field of transportation engineering. Finally, it identifies many open research subjects in transportation in which the use of RL seems to be promising. © 2003 NRC Canada.