DEVELOPMENT OF A GAME ENGINE-BASED TOOL FOR TRAFFIC MICROSIMULATION USING MOBILITY BEHAVIOURS

Abstract

Gaming tools gained significant interest among transportation modellers for mobility analysis, visualization and scenario testing. Commercial simulation software programs were used for running traffic microsimulations for years. They serve as traditional tools for both macroscopic and microscopic analyses. However, they lack flexibility when testing new methods or scenarios. This thesis presents a framework to create an open-ended 3D traffic microsimulation tool. It is achieved by incorporating a virtual 3D environment and integrating mobility behaviour models of vehicles and pedestrians into a game engine, thus creating a robust simulation tool. The tool is tested by simulating pedestrian scenarios before and after pandemic situations to analyze the percentage of violations with the help of a social force model. The tool’s flexibility is then tested by using it to evaluate performance differences in various kinds of traffic consisting of autonomous and human-driven vehicles.