Risk Assessment Method for Connected and Autonomous Vehicles Based on Traffic Factor Safety State Network and Risk Field Theory

The driving risk of a vehicle is influenced not only by its own state, but also by the driving environment. With the development of smart connected technologies, vehicles can acquire more information about the environment. However, traditional driving risk assessment indicators only consider the impact of individual traffic factors and are unable to assess driving risk comprehensively. Moreover, they fail to evaluate the overall safety of mixed traffic flow from a macroscopic perspective. In this study, we model various types of risks based on the risk field theory and analyze the environmental impact factors of road sections based on the traffic factor safety state network theory, thereby proposing a microscopic vehicle risk field indicator VRFI and a novel macroscopic road risk field indicator RRFI. In the experiment part, we first verify the rationality of the microscopic vehicle risk assessment indicator VRFI through car-following and lane-changing scenarios, and then validate the effectiveness of the macroscopic road risk monitoring indicator RRFI using real-world traffic data and simulation techniques. Finally, we analyze the overall risk of traffic flow under different traffic states. This study provides an innovative method for microscopic risk assessment of connected and autonomous vehicles and macroscopic risk monitoring of mixed traffic flow.