Evaluating the Effects of Variable Speed Limit on Traffic Safety under Adverse Weather Conditions Based on Trajectory Data

Adverse weather, like haze, rain, etc., can decrease drivers’ visibility and increase vehicles’ stopping distances, thus augmenting the driving safety risks, leading to potential accidents. Variable speed limit (VSL) is a primary traffic control strategy to enhance road traffic safety. This research proposes a safety risk assessment method to evaluate the effects of VSL on traffic safety under adverse weather conditions based on trajectory data, considering conflict types. Surrogate safety measures and a logistic regression model using aggregated traffic flow data and crash data are utilized to identify the potential longitudinal and lateral conflicts. Simulation scenarios with/without VSL under different compliance rates are built for normal/adverse weather conditions to evaluate the safety risk with VSL. Furthermore, an XG Boost-based risk prediction model is developed. The results show VSL strategy has effective capability to reduce traffic safety risks under adverse weather on highways, especially when the compliance rate is high.