Characteristics of Mixed Traffic Flow at Accident Bottlenecks with Connected and Autonomous Vehicle Platoons

This study investigates the impact of connected and autonomous vehicles (CAVs) and platoons on mixed traffic flow at accident bottlenecks. A cellular automata (CA) model is proposed, incorporating car-following and lane-changing models based on platoon communication range and formation mechanisms, specifically designed to simulate traffic flow characteristics at accident bottlenecks. The findings indicate that the impact of accident bottlenecks on traffic flow correlates with traffic density, with accident duration exerting a more significant influence than the length of the warning zone. As accident duration increases, the critical density range of congested flow at the bottleneck expands to 65 vehicles/km. Under cooperative adaptive cruise control (CACC) mode, reduced safety distances improve traffic flow. Although expanding communication range enhances traffic efficiency, the marginal benefit diminishes beyond a range of 80 meters. Increasing the proportion of CAVs significantly boosts traffic flow, alleviates congestion, and enlarges both the number and scale of platoons. When the CAV proportion reaches 100%, traffic flow increases linearly with density, while average speed remains at a high level.