Connected and automated vehicle (CAV) technologies promise a substantial decrease in traffic accidents and traffic jams, and bring new opportunities for improving vehicle’s fuel economy. However, testing autonomous vehicles in a real world traffic environment is costly, and covering all corner cases is nearly impossible. Furthermore, it is very challenging to create a controlled real traffic environment that vehicle tests can be conducted repeatedly and compared fairly. With the capability of allowing testing more scenarios than those that would be possible with real world testing, simulations are deemed safer, more efficient, and more cost-effective. In this work, a full-scale simulation platform was developed to simulate the infrastructure, traffic, vehicle, powertrain, and their interactions. It is used as an effective tool to facilitate control algorithm development for improving CAV’s fuel economy in real world driving scenarios. The simulator integrates a 3D traffic model with a high-fidelity vehicle model using a modular architecture, which supports hardware-in-the-loop (HIL) and vehicle-in-the-loop (VIL) testing. Sensor models, perception module and manual/autonomous driving modules can be customized for the test vehicle. Smart and scalable traffic scenarios can be easily generated on the road network using statistically representative trip data. The design, implementation, and usage of the simulator will be described.
Virtual Traffic Simulator for Connected and Automated Vehicles
Sae Technical Papers
WCX SAE World Congress Experience ; 2019
2019-04-02
Conference paper
English
Virtual Traffic Simulator for Connected and Automated Vehicles
British Library Conference Proceedings | 2019
|Virtual Rings on Highways: Traffic Control by Connected Automated Vehicles
Springer Verlag | 2022
|Safety-critical traffic control by connected automated vehicles
Elsevier | 2023
|