Understanding the Safety Performance of Traffic Flow Mixed with Connected and Automated Vehicles in Snowy Weather Conditions Using Domain Crash Risk Entropy

The improvement of traffic flow safety under snowy weather conditions has long been a formidable challenge for transportation systems. Existing countermeasures can only passively control and alleviate traffic crash risks arising from the snowy weather conditions. The advent of connected and automated vehicles (CAVs) holds promise for rendering traffic flow safer and more stable. However, the safety of traffic flow mixed with CAVs under snowy weather conditions remains poorly understood. The objective of this study is to investigate the safety of traffic flow under snowy weather conditions, considering varying market penetration rates of CAVs. The study conducts a thorough analysis of operational data obtained from an intersection located in Harbin, China, which encompasses traffic flow parameters during distinct weather conditions such as summer sunny, winter sunny, light snow, and moderate snow. Furthermore, the study provides insights into the influence of snowy weather on the intersection traffic flow efficiency. Additionally, a novel concept called the domain crash risk entropy is introduced. This concept represents the uncertainty associated with traffic flow speed and acceleration distributions. Simulations are conducted to examine the safety of mixed traffic flow in snowy weather conditions. The results show that CAVs have the potential to enhance traffic flow operation efficiency, stabilize acceleration distribution, and reduce the uncertainty of traffic flow speed, thereby ensuring traffic flow safety.