A Trajectory Data‐Driven Study on the Evolution Mechanism and Control Strategies of Lane‐Changing Behavior in Intersection Areas of Expressways

Vehicle interactions in weaving sections are relatively frequent and complex, posing significant challenges to traffic congestion management and safety. Trajectory data‐driven driving behavior analysis can effectively reveal differences in driving behaviors. Therefore, in accordance with the research requirements, this study selected the CitySim dataset as the foundation after comparison and utilized intelligent algorithms to extract 1349 lane‐changing samples from a specific weaving section within the dataset for analyzing the lane‐changing behavior characteristics of vehicles in weaving areas. After analyzing the sample data using traffic flow theory and statistical theory, the following results were obtained: the speeds increase upon entering and decrease upon exiting weaving zones, while headway distances consistently grow. Vehicles in the inner lanes exhibit smoother transitions and higher speeds, while lane‐changing speeds range from 10 to 55 km/h (median: 29 km/h) and durations vary from 2 to 18 s (median: 8.5 s). Statistical analyses highlight significant behavioral differences based on lane and direction. Vehicles on entrance ramps demonstrate higher speeds, longer durations, and larger headways than those on exit ramps. Furthermore, right‐lane changes are associated with lower speeds and shorter durations compared with left‐lane changes. Based on these findings, the study proposes targeted traffic management strategies, including ramp flow control and optimized road markings, to enhance safety and efficiency in weaving areas. This research provides actionable insights for traffic management and road design in the expressway weaving areas.