Influence of Trap Length, Snapshots, Traffic Volume and Composition on Density Estimation on a Highway

A thorough understanding and correct measurement of traffic density is essential for proper management of traffic and planning of transportation infrastructure and policies. Traditionally, density is estimated by dividing measured flow to measured space mean speed using fundamental traffic equation. Other methods of density estimation available in literature either give highly erroneous results in mixed traffic flow (for example highway capacity manual method) or require instrumented highways for their application (like n-t method). This necessitates developing other method to estimate density which can be used under heterogeneous traffic situation as prevalent on non-instrumented roads of developing countries such as India. This paper proposes a new method, named multi snap method which is a modification to Highway Capacity Manual (HCM) method; to estimate density. This method involves taking a number of snap shots of the roadway section under study during the study period so that the error in density estimation reduces to some acceptable level. It is found that results are sensitive to a number of factors like trap length, traffic volume and composition which are further explored by using a simulation program VISSIM. It is observed that the error in density estimation decreases with increase in the trap length and traffic volume. Minimum trap length required to get the density as good as in one kilometre of the highway also reduces with increase in the traffic volume. Increase in share of two-wheelers in the traffic stream increases the error in density estimation whereas heavy vehicles in the stream reduce the error.