Using Reinforcement Learning to Find the Shortest Path between two Locations on the Public Roadways

ABSTRACT It goes without saying that population growth, increasing number of vehicles, and unprecedented air pollution in recent years have led to traditional urban transportation planning systems no longer be as efficient as possible. The artificial intelligence provides solutions for many problems and one of its methods is the reinforcement learning (RL). In this article a method has been proposed based on RL to improve the quality of the transportation services, which in turn decreases the traffic jam and air pollution. The proposed method finds the shortest route between source and destination points and avoids routes with traffic congestion which both lead to decrease in travel time, and decline fossil fuel and energy consumption.

Keywords:

Keywords: Reinforcement Learning (RL) , Transportation , Single-Agent reinforcement learning , Q-Learning