Optimization of Autonomous Vehicle Safe Passage at Intersections Based on Crossing Risk Degree

In the context of autonomous driving, ensuring safe passage at intersections is of significant importance. An effective method is necessary to optimize the passage rights of autonomous vehicles at intersections to enhance traffic safety and operational efficiency. This paper proposes an analytical model for assigning the right-of-way to autonomous vehicles approaching intersections from different directions. Assuming that fully autonomous vehicles equipped with advanced Vehicle-to-Everything (V2X) communication and real-time data processing can utilize gaps to proceed at unsignalized intersections in the future, the Crossing Risk Degree (CRD) indicator is introduced for safety assessment. A higher CRD value indicates a higher crossing risk. CRD is defined as the product of the kinetic energy loss from collisions between vehicles in the priority and conflicting fleets, and the probability of conflict between these two fleets. By comparing CRD values, the passage priority of vehicles at intersection entrances can be determined, ensuring efficient passage and reduced conflict risks. SUMO microsimulation modeling is employed to compare the proposed traffic optimization method with fixed signal control strategies. The simulation results indicate that under a traffic demand of 1200 vehicles per hour, the proposed method reduces the average delay per entry approach by approximately 20 s and decreases fuel consumption by about 50% compared to fixed-time signal control strategies. In addition, carbon emissions are significantly reduced. The findings provide critical insights for developing intersection safety management policies, including the establishment of CRD-based priority systems and real-time traffic monitoring frameworks to enhance urban traffic safety, symmetry, and efficiency.