Freeway Gap Acceptance Behaviors Based on Vehicle Trajectory Analysis

Gap acceptance forms a significant component of lane change modeling. Most studies on gap acceptance focus either on intersections or on merging areas. There is relatively little research on gap acceptance behaviors during lane changes on freeways. This paper analyzed gap acceptance behaviors on freeway segments by conducting a detailed analysis of the vehicle trajectory data obtained through Next Generation SIMulation (NGSIM) program. The paper classified accepted gaps from the vehicle trajectory data into mandatory lane change (MLC) and discretionary lane change (DLC) gaps, and conducted a detailed regression analysis by studying 24 explanatory variables grouped into six categories. The analysis shows that MLC maneuvers require greater gaps than DLC maneuvers. Also, if multiple lanes need to be traversed for a MLC, drivers are more inclined to accept smaller gaps. Presence of Heavy vehicles as subject or lag vehicles also tend to increase accepted gaps for both MLC and DLC maneuvers, while front and lead vehicle types do not have significant effect. Regression analysis revealed that the variance of observed variables, including MLC-specific variables, vehicle types, speed, relative speed, acceleration, and relative acceleration, can explain the variance of accepted gaps about 49.5% and 38.9% for MLC and DLC, respectively. This suggests that some other unobserved factors, such as gender, age, trip purpose, etc., may also play a role in gap acceptance when conducting lane changes on freeways. Overall, this paper enriches the understanding of gap acceptance behaviors when driving on freeways.