Abstract:
Traffic networks which includes many kinds of networks like highways, streets etc., provide convenient and economical conveyance of passengers and goods. The basic activity in transportation being a trip, which is defined by destination and origin, arrival and departure times and the route taken for travel. Different trips interact with each other on the network to produce complex pattern of traffic flows. Since traffic conditions in major cities and urban conglomerations are becoming increasingly congested, affecting the overall operational efficiency of networks as well as the cost of travel of each individual trip, modelling of traffic flow is being seen as essential rather than secondary process in traffic engineering and the policy making process in transport sector.
Traffic flow is one such phenomenon which is highly difficult to model mathematically due to its extreme complexity. The most basic traffic scenario is a one dimensional road with one-way traffic. This is a simple scenario to model as we have eliminated various factors.
There are three main approaches taken to model traffic flow; microscopic approaches, macroscopic approaches and mesoscopic approach. Microscopic models map traffic flow as a set of individual vehicles, while macroscopic models map traffic flow as fluid flow where each vehicle is analogous to a molecule of fluid. Mesoscopic models describe vehicle behaviour in aggregate terms such as in probability distributions. They essentially cover the ground in between a macroscopic and a microscopic model.
In this report, the main emphasis is on macroscopic models. Macroscopic models place more emphasis on traffic flow as a continuum versus a collection of individual vehicles. Continuum traffic flow modelling generally uses a macroscopic perspective, although microscopic principles can be incorporated into continuum models
