Modelling passenger car equivalency at an urban midblock using stream speed as measure of equivalence

The effect of traffic volume and its composition on Passenger Car Equivalency (PCE) of different vehicle types in a mixed traffic stream is investigated taking an urban mid-block section as the case study. The reduction in stream speed caused by marginal increment in traffic volume by a vehicle type is compared with that of caused by an old technology car, for the estimation of PCE of that vehicle type. A Neural Network (NN) approach is explored for capturing the underlying non-linear effects of traffic volume and its composition level on the stream speed. It is found that PCE of a vehicle type varies in a non-linear manner with total traffic volume and compositional share of that vehicle type in the traffic stream. The speed model using NN technique alone could establish the variation of PCE with vehicle type, traffic volume and its composition.