Modeling Freeway Diverging Behavior on Deceleration Lanes

Abstract

Typically, deceleration speed-change lanes (SCLs) and exit ramp terminals on freeways are designed to allow vehicles to depart the freeway safely and efficiently. Freeway diverge areas with adequate SCL lengths should enable the exiting drivers to diverge off the freeway through traffic, decelerate to the desired ramp speed, and exit the freeway safely and comfortably. Safety problems can be expected if drivers are forced to reduce speed on the main traffic lanes or to decelerate at a very high rate. In Canada and the United States, the current design guides suggest that operation and safety could be enhanced on freeway SCLs by increasing their length. In this study, driver speed behavior at freeway diverge areas was examined by using data collected on 13 exit ramp terminals on Highway 417 in the city of Ottawa, Canada. The speed, geometric, and traffic data were collected and employed to model 85th percentile speeds and deceleration rates using linear regression analysis. The modeling attempts integrated in this study resulted in eight statistically significant predictive models at the 5% level of significance.