Safety performance evaluation of freeway merging areas under autonomous vehicles environment using a co-simulation...

Area of Science:

• Traffic Engineering and Transportation Science
• Intelligent Transportation Systems (ITS)
• Autonomous Vehicle Safety

Background:

• Freeway merging areas are critical bottlenecks for traffic flow and prone to incidents due to human driver errors.
• Evaluating the safety impact of autonomous vehicles (AVs) is hindered by a lack of real-world data and limitations in existing simulation tools.
• Previous studies often rely on single simulators, failing to accurately model AV behavior and complex traffic scenarios.

Purpose of the Study:

• To address the challenges in evaluating AV safety by developing a high-fidelity co-simulation platform.
• To customize car-following models for AVs and assess their performance in freeway merging scenarios.
• To investigate the safety and comfort impacts of both single and multiple AVs in mixed traffic environments.

Main Methods:

• Developed a software-in-the-loop co-simulation platform integrating PreScan (environmental perception), Matlab (AV decision-making/control), and Vissim (traffic flow).
• Customized various car-following models for AVs to accurately represent their behavior in merging conflicts.
• Conducted experiments on a real freeway merging scenario, analyzing safety (Time-to-Collision, TTC) and comfort (jerk) for single and multiple AVs.

Main Results:

• Microscopic analysis of single AVs revealed the performance of different car-following models during merging conflicts.
• Macroscopic risk assessment demonstrated substantial reductions in traffic conflicts and risks with increasing AV market penetration.
• The co-simulation platform successfully reproduced mixed traffic scenarios with AVs, enabling comprehensive safety evaluation.

Conclusions:

• Autonomous vehicles offer significant potential to enhance freeway safety, particularly in bottleneck areas like merges.
• The developed co-simulation framework provides a robust tool for evaluating AV safety performance in realistic traffic conditions.
• Increased adoption rates of AVs are projected to lead to substantial improvements in overall traffic safety and risk mitigation.