An Elegant and Computationally Efficient Approach for Heterogeneous Traffic Modelling Using Agent Based Simulation

Abstract

For developing nations, where mixed traffic conditions prevail, it becomes necessary to develop a heterogeneous traffic flow model which is able to model all vehicle types of different static and dynamic characteristics and equally efficient in computational performance. For this purpose, in this study, an agent based framework is used which uses a queue model in its mobility simulation. In first step, queue model is enhanced by adding different vehicle types with different maximum speeds and sizes. Furthermore, traditional FIFO (First-In-First-Out) approach of queue model is modified to a more realistic modified queue model in which faster vehicles can pass slower vehicles. The enhancements are discussed with their fundamental diagrams for traffic flow and spatio-temporal diagrams.