Modeling vehicle acceleration-deceleration behavior during merge maneuvers

Freeway merge maneuvers are complex and may involve a lane change, continuous acceleration, possibly deceleration, and finally a gap acceptance. This paper presents a methodology to model ramp vehicle acceleration-deceleration behavior at freeway entrance ramps. Primary data capturing a wide ranee of information were collected using video tape methods and manually reduced through image processing techniques. Ramp vehicle merging position in conjunction with freeway and ramp flow levels were analyzed for both parallel and taper type entrance ramps. Where vehicle trajectory data were available, merging positions with respect to ramp vehicle speed as well as relative speed and time gap between a ramp vehicle and freeway vehicles were examined. Results show that the merge position has a statistically insignificant effect upon individual traffic parameters. Methodology presented in this paper used the stimulus-response equation as a fundamental rule to model ramp vehicle acceleration-deceleration behavior. The time-base car-following model was modified to a nonlinear distance-base stimulus-response equation to capture the characteristics of the collected traffic data. Calibration of the proposed nonlinear model from a pilot study demonstrated satisfactory results and indicated that a time lag before ramp drivers respond to stimuli does exist. Effects of surrounding freeway vehicles on ramp vehicle acceleration-deceleration behavior were found to be significant.