Connected vehicle enabled hierarchical anomaly behavior management system for city-level networks

Abstract

Drivers who are distracted cannot operate their vehicles appropriately, which leads to error-prone behavior on the roads. This behavior increases the risk of collisions for both themselves and surrounding vehicles, making it urgent to manage anomalous vehicles with distracted drivers and mitigate their impacts on driving safety. To address this problem, this paper presents an anomaly behavior management system that leverages connected vehicles to improve the safety performance for both individual vehicles and the whole network. The proposed system integrates a hierarchical architecture that reduces the risk of collisions caused by anomalous vehicles in large-scale road networks. Connected vehicles monitor anomalous vehicles and estimate speed and lane-changing instructions to avoid dangerous behaviors. The benefits of the proposed system are evaluated using microscopic traffic simulation, which shows a reduction in the risk of collisions and improved mobility for both connected vehicles and the entire network. The paper also conducts a sensitivity analysis of the market penetration rates of connected vehicles and traffic demand levels to understand the system’s reliability at different development stages of connected vehicles and traffic congestion.