Simulation of Autonomous Vehicles Based on Wiedemann's Car Following Model in PTV Vissim

A lot of research deals with capacity effects of autonomous driving, but little is known about the exact driving behavior of autonomous vehicles. Most simulations are based on assumptions regarding the following behavior. Different technical components, that might influence decisively the car following behavior in future vehicles, further complicate the topic. Therefore, there is no common agreement or recommendations on parameter settings when modeling autonomous vehicles in microsimulation software. The authors analyzed data of two autonomous vehicles driven in a real-world scenario and compared their following behavior to simulations based on the Wiedemann car following model. They derived the standstill distances and the headway policy of autonomous vehicles from the data. The tests comprised driving in autonomous mode with and without communication between two following vehicles. The simulations show that the behavior of autonomous vehicles communicating with the leading vehicle are reproduced well in Vissim. Problems remain when simulating autonomous vehicles that do not communicate with their leader.

• Authors:
  • Zeidler, Verena
  • Buck, H Sebastian
  • Kautzsch, Lukas
  • Vortisch, Peter
  • Weyland, Claude Marie
• Conference:
  • Transportation Research Board 98th Annual Meeting
  • Location: Washington DC, United States
  • Date: 2019-1-13 to 2019-1-17
• Date: 2019

Language

• English

Media Info

• Media Type: Digital/other
• Features: Figures; References; Tables;
• Pagination: 12p

Subject/Index Terms

• TRT Terms: Autonomous vehicles; Car following; Computer models; Connected vehicles; Microscopic traffic flow
• Identifier Terms: VISSIM (Computer model)
• Subject Areas: Highways; Operations and Traffic Management; Planning and Forecasting;

Filing Info

• Accession Number: 01711091
• Record Type: Publication
• Report/Paper Numbers: 19-02482
• Files: TRIS, TRB, ATRI
• Created Date: Jul 17 2019 9:49AM