Development and validation of a simplified microscopic multilane model for traffic flow simulation.

摘要

Existing microscopic simulation models incorporate detailed car following (CF) and lane changing (LC) models that require higher computational resources. This research presents the development and validation of a simplified microscopic model, CELLSIM, for simulation of traffic at the highway network. The model formulation uses concepts of cellular automata (CA) and CF models, but is more detailed than CA models and has realistic acceleration and deceleration behavior. A dual-regime acceleration model is developed that requires minimum calculation, is easier to calibrate and provides a good fit to the field data. A simplified CF logic is developed, which determines the follower's acceleration, deceleration, and coasting behavior based on a 3 by 3 matrix. The matrix compares the available and desired space gaps, and speed of the follower and the leader to determine the follower's action. Desired space gap is a product of follower's preferred time headway (TP) and speed, where TP represents the choice headway of individual drivers in steady state CF and can be found from the field data. Collision avoidance logic is also incorporated, and CELLSIM performs well when tested for normal and emergency disturbances for a platoon of vehicles.; Simplified LC models for discretionary and mandatory LC are developed. The LC models are calibrated for a weaving section and are used in conjunction with the CF logic to develop a multilane model. Discretionary LC logic considers a driver's dissatisfaction with his/her current speed and feasibility of speed gain in the target lane. The mandatory LC logic is based on collision avoidance and the lane changer adjusts its speed by decelerating or coasting to move to the target lane. The LC models require fewer calibration parameters compared to existing models. The simplified approach makes CELLSIM suitable for main line control of traffic in real time.; CELLSIM performs well in congested and non-congested traffic conditions and has been validated comprehensively at the macroscopic and microscopic levels. Comparison of field data and CELLSIM for trajectories, average speed, density and volume show very close agreement. Statistical comparison of macroscopic parameters with other CF models indicates that CELLSIM performs as well as detailed models. CELLSIM is also able to replicate the hysteresis effect.