HEADWAY MODELLING UNDER MIXED TRAFFIC ON URBAN ROADS

Abstract

Headway modelling' has become integral part of study for the traffic flow system in many advanced countries. Traffic flow is an extremely complex phenomenon and its complete understanding is quite difficult. However, headway can help to some extent for understanding nature of traffic flow. The main object of headway modelling is to identify the headway impacts on arrival pattern, driver behaviour, accident study and traffic flow control at intersection. As a consequence of the importance of headways as traffic flow descriptor, considerable effort has been expended for the purpose of developing a sound, general purpose headway model. The case for introducing such tool in heterogeneous nature of traffic stream is even stronger, as many new traffic models are being planned and/or implemented at state and central level. In the present study, an effort has been made to model headway data collected at a few sections of urban roads in Delhi. Among various types of mathematical distributions hyperlang model was found to be the best descriptor of headway under mixed traffic condition. The model parameters were found to be related with traffic volume on a road. The generalised hyperlang model suggested in this study is applicable to traffic volume range of 900 vph to 1600 vph per lane.