Gap Acceptance Behavior in Mandatory Lane Changes under Congested and Uncongested Traffic on a Multi Lane Freeway


Gap acceptance is an integral part of lane change models in microscopic traffic simulation. Offered gap is considered as an accepted gap when it is greater than or equal to a minimum value. The minimum value of accepted gap for the intended lane change maneuver is generally defined as the critical gap. Critical gap is a function of accepted and rejected gaps and depends on the driver behavior parameters like aggressiveness, urgency, impatience. Drivers behave differently under diverse traffic, geometric, and environmental conditions. Similarly, the same driver can behave differently under diverse conditions. Due to the stochastic driver behavior, modeling critical gaps mathematically to get realistic results from microscopic traffic simulation models is imperative. This paper presents results from a study of critical gaps for mandatory lane changes using data collected by NGSIM project on I80 California (a multi-lane freeway) in congested and uncongested traffic flow conditions. Both leading and trailing critical gaps have been estimated separately using deterministic and stochastic methods. Comparison of results between these methods is presented. Trailing critical gaps are more sensitive compared to leading critical gaps. Additionally, critical gaps are presented as a function of lane change location, upstream of an off-ramp. Under uncongested traffic flow, impulse for lane change is observed between 800 and 1000 feet upstream of off-ramp. Distribution of accepted gaps for mandatory lane changes are determined and based on the results Gamma distribution is proposed for accepted gaps.


Civil, Architectural and Environmental Engineering

Keywords and Phrases

Gap Acceptance; Lane Changing; Multilane Highways; Operations and Traffic Control; Traffic Congestion; Traffic Simulation; Vehicle Mix; Traffic Flow

Document Type

Article - Conference proceedings

Document Version


File Type





© 2007 National Academy of Sciences, All rights reserved.