A VR Approach to Simulate the Interactions of Pedestrians and Vehicles in a Pedestrian Crossing

Abstract

The increase in the number of vehicles in cities creates lots of critical problems to the society. The environmental impact being the most signi cant tragedy, the growth in road accidents has also become a major problem. Pedestrian crashes hold a signi cant score among the road accidents. In the process of nding solutions and evaluating proposed solutions to develop better pedestrian facilities, there exists the need of pedestrian crossing analysis tools and safety tools to conduct pedestrian studies. The motivation for this research is to build a reusable model that can be employed in rapid development of various simulation systems. The goal of this research is to develop a new model that is able to simulate the real interactions of pedestrians and vehicles in a pedestrian crossing. First the existing techniques for autonomous pedestrian movement and vehicle movement is investigated and decided on an approach to be applied for pedestrian behaviors. Vehicle behavior has been decided to model in a novel approach. And this research suggests a novel model of pedestrian vehicle interactions by means of behavioral algorithms. The proposed system architecture comprises of four component units to control the interactions of pedestrians and vehicles in a pedestrian crossing. The pedestrian behaviors are modeled by psychology related behavioral algorithms and the vehicle behaviors involve the characteristics of data structures. A prototype simulation is implemented on the simulation engine VBS and the experiments prove that the proposed novel model is e ective to use.