Reinforcement Learning for Vehicle Route Optimization in SUMO

Urban traffic control becomes a major topic for urban development lately as the growing number of vehicles in the transportation network. Recent advances in reinforcement learning methodologies have shown highly potential results in solving complex traffic control problem with multi-dimensional states and actions. It offers an opportunity to build a sustainable and resilient urban transport network for a variety of objects, such as minimizing the fuel consumption or improving the safety of roadway. Inspired by this promising idea, this paper presents an experience how to apply reinforcement learning method to optimize the route of a single vehicle in a network. This experience uses an open-source simulator SUMO to simulate the traffic. It shows promising result in finding the best route and avoiding the congestion path.