Probabilistic Model for Design of Freeway Acceleration Speed-Change Lanes

Abstract

A speed-change lane (SCL) is an auxiliary lane added to the traveled way for the acceleration and deceleration of vehicles entering or leaving a roadway. When the length of an acceleration SCL is adequate, drivers are able to accelerate comfortably from the speed at entrance to a speed appropriate to the road, find a gap in the traffic flow, and merge in a safe and secure manner. The length of an SCL is currently determined in terms of the ramp design speed, the freeway design speed, and the acceleration rate. Embedded in these values are assumptions for the operating speed at the entrance and merging points. This study examined a probabilistic approach instead of such a deterministic approach. The main benefit of a probabilistic approach is that traffic flow characteristics are assumed to be stochastic; therefore, the outcome of a probabilistic methodology is a distribution of drivers’ acceleration distance on the SCL. The reliability-based analysis enables designers to select a specific percentile value of this distribution as a design length that better matches a certain situation and avoids unnecessary extra construction costs. This paper presents analytical and simulation models for the application of the reliability approach, with all parameters based on recently collected field data. Even though the presented model should be superior to the deterministic model adopted in current design guides, additional enhancements are recommended for a full reliability-based, safety-explicit design model.