Effects of AV control strategies on mixed traffic flow under highway incidents with service area diversion

Abstract

This study investigates mixed traffic flow under highway incident conditions with the aim of evaluating the effectiveness of automated vehicle (AV) control and service area-based traffic management. A cellular automaton model is developed that integrates AV proactive lane avoidance, probabilistic service area diversion, and dynamic release strategies based on real-time traffic conditions at the bottleneck. Simulation experiments are conducted to analyze the effects of AV penetration rate, vehicle avoidance distance, and service area capacity under different incident locations. The results show that the improvement in system performance requires a sufficiently high AV penetration rate. There exists an appropriate range of AV avoidance distance that leads to better system performance under incident conditions. In addition, the effectiveness of service area capacity depends strongly on its spatial relationship with the bottleneck. These findings provide theoretical and practical implications for the design of AV control strategies and highlight the importance of their interaction with spatial buffering under incident conditions.