Modelling communication-enabled traffic interactions.

Area of Science:

• Autonomous Systems
• Human-Computer Interaction
• Computational Social Science

Background:

• Autonomous vehicles (AVs) face challenges interacting with human drivers, particularly in complex scenarios like highway merging.
• Current computational models often overlook driver communication and assume one-sided responses, limiting their accuracy in real-world interactions.

Purpose of the Study:

• To develop a novel computational framework for modeling interactive driver behavior.
• To address limitations in existing models by incorporating explicit driver communication and bounded rationality.

Main Methods:

• Proposed a joint interactive system model, moving beyond isolated driver simulations.
• Integrated explicit communication channels and bounded rationality into the driver behavior model.
• Tested the framework in simplified highway merging and car-following scenarios.

Main Results:

• The model generated plausible interactive behaviors, including aggressive and conservative merging strategies.
• Human-like gap-keeping behavior emerged in car-following simulations, driven by risk perception.
• The framework demonstrated the potential for modeling complex, two-way driver interactions.

Conclusions:

• The developed framework offers a promising approach for accurate interaction modeling.
• This research can significantly contribute to the development of interaction-aware autonomous vehicles.
• Explicitly modeling communication and bounded rationality is key to realistic driver behavior simulation.