Traffic Risk Modelling and Analysis under Airport-Like Simulation Environment

Traffic safety plays a crucial role in the development of autonomous vehicles which attracts significant attention in the community. It is a challenge task to ensure autonomous vehicle safety under varied traffic environment interference, especially for airport-like closed-loop conditions. To that aim, we analyze autonomous vehicle safety at typical roadway conditions and traffic state constraints (e.g., car-following state at different speed distributions) by simulating the airport-like traffic conditions. The experimental results suggest that traffic collision risk is in a positive relationship with the speed difference and distance among adjacent vehicles. More specifically, the autonomous vehicle may collide with neighbors when the time to collision (TTC) indicator is lower than 4 s, and vice versa. The research findings can help both research community and practioners obtain additional information for improving traffic safety for autonomous vehicles.