Modelling multi-class disordered traffic streams using traversable distance: a concept analogous to fluid permeability

Abstract

In multi-class disordered traffic conditions, the presence of numerous classes of vehicles having different size and operational characteristics invalidates the behaviour of lane discipline. Each vehicle traverses through a series of both viable as well as accessible gaps created by other vehicles, thus defining a permeable medium. In this paper, a plausible way to incorporate the disordered behaviour on macroscopic traffic stream modelling is discussed. A new metric termed as traversable distance has been formulated for characterising this behaviour, based on field observations. Based on class-specific traversable distances that correspond to a set of incommensurate density measures, equilibrium speed functions have been redefined. Disordered stream behaviour significantly affects the scatter in the equilibrium speed–density relationship which is a key input to macroscopic traffic flow models. Results from the modified speed functions based on traversable distance proved that the proposed methodology could considerably explain the scatter observed in the fundamental diagram.