Citation

Mukhopadhyay D, Sarkar P. Indian Highw. 1992; 20(10): 7-16.

Copyright

(Copyright © 1992, Indian Roads Congress)

DOI

unavailable

PMID

unavailable

Abstract

This paper outlines some well-known models for estimating turning traffic at an intersection with N arms, applies them to two examples of urban traffic in India, and evaluates the model performances by comparing observed and simulated results. In non-iterative growth factor models, observed traffic movements are extrapolated to estimate future traffic patterns. These models include: (1) the uniform growth model, where the same ratio of future and existing traffic is applied to each turning movement; and (2) the split model, where existing turning probabilities are multiplied by future inbound and outbound flows to obtain future turning movements. Different iterative models, which simulate observed turning movements and are basically growth factor methods, include: (1) the Fratar model; (2) the Furness model; and (3) the Krutihoff model. The models were tested using turning traffic flow data, collected during morning and evening peaks in 1975 and 1987 from two road intersections with mixed traffic conditions, one in Delhi and one in Meerut. Entropy formulae were used to compare the observed and simulated results, although the chi- square test could have been used. The Furness model was found to produce very reliable estimates when no large change in turning flow patterns is foreseen; otherwise, the Krutihoff model was more dependable.