Agent-based microscopic pedestrian interaction with intelligent vehicles in shared space

Abstract

Looking out for pedestrians has long been one of the most important issues for intelligent vehicles. Sometimes, intelligent vehicles have to cope with a large crowd of pedestrians. This is extremely common in shared spaces such as campus, shopping mall, or transportation station. In this paper, a vehicle-pedestrian interaction simulator is introduced to help intelligent vehicles, especially automated vehicles in The Smart Mobile Operation: OSU Transportation Hub (SMOOTH), achieve better decision making and local path planning. This simulator integrates different characteristics of pedestrians, and is capable of easily configuring and simulating different scenarios. Fundamental simulations have been completed to verify the simulator effectiveness for a pilot four-passenger golf cart. The results show that the simulator can effectively reflect the interaction behavior between vehicles and pedestrians.