Jorge Laval

Lagrangian Resolution of the Multilane Hybrid Traffic Flow Model

It has been recently shown that most traffic instabilities and complex traffic behavior can be traced down to the effects that lane
changes have in traffic streams. Up to now the effects of lane changes have been modeled successfully using the Multilane Hybrid (MH)
model; ie, a continuum kinematic wave stream in each lane occasionally interrupted by lane changes, which are treated as discrete
particles with bounded accelerations. A macroscopic theory of vehicle lane-changing is used in the MH model, where lane changes take
place according to a stochastic process whose mean depends on lane-specific macroscopic quantities.
In this talk the Lagrangian resolution of the MH model will be presented, focusing on the appropriate representation of the macroscopic
 lane changing model. The resulting lane-changing logic becomes very simple compared to existing microscopic models, and requires only
one extra parameter. The model is shown to agree nicely with empirical data.