A Study on the Methodology to Develop Virtual Drive Environment for Autonomous Driving Evaluation

Abstract

It takes many hours and restrictions once AD(Autonomous Driving) evaluation based on real tests. This paper presents a methodology for development of virtual driving environment that can replace the real vehicle test. When developing a virtual driving environment, it is important to develop the same virtual element model (Road, Vehicle model, etc.) as the real world. So the high-occupancy BRT (Bus Rapid Transit) bus route in Cheongna zone was modelled using the MMS(Mobile Mapping System) as the openDRIVE format which is the ASAM(Association for Standardization of Automation and Measuring Systems) road standard. In addition, we develop a vehicle model that simulates the dynamic performance of BRT based on Modelica language. Finally, we develop an interface module that integrates the virtual environment, the vehicle model, and the driver model. In conclusion, this paper present virtual test drive platform for AD Evaluation.