Simulating Lane-changing Dynamics towards Lane-flow Equilibrium Based on Multi-lane First Order Traffic Flow Model

Abstract

It is well known that under the condition of high traffic volume lane-flow distribution becomes unbalanced; more traffic tends to use a median lane rather than a middle and outer lane, which causes the deterioration of traffic capacity at bottleneck sections. As intensive development of ITS, active and dynamic lane management has been practically implemented. By employing the technology of ITS, balancing lane-flow distribution is one of the feasible solutions to increase the throughput of bottleneck flow. However, the mechanism of the unbalanced usage among lanes is still unclear and traffic flow models enabling online and network-wide evaluation of dynamic and strategic lane management have not been proposed due to the lack of the method to computing lane-based traffic flow including lane-changing dynamics. This paper developed the multi-lane first order traffic flow model, which depicts the dynamics of lane-changing. It is assumed that each vehicle changes the lane to improve its utility or decrease its disutility, and also that the equilibrium of lane flow distribution is achieved as the condition of stochastic user equilibrium (SUE), where all drivers believe that they cannot improve their utility by changing the lanes. Thus, in the model, lane-changes are represented as the dynamics towards lane-flow equilibrium. As a result of simulating multilane traffic on freeway without any merging and diverging, it is revealed that the proposed model can represent the equilibrium curve of lane-flow distribution, and depict the propagation of traffic congestion at a lane-drop bottleneck section.