Modelling drivers’ behaviour within the dilemma zone at traffic signal junctions

The current study introduces a newly developed, calibrated, and validated micro-simulation model for predicting drivers’ decisions following the onset of an amber traffic light signal under the effect of the dilemma zone where a driver can neither stop safely nor cross and clear the junction before the onset of red. The purpose of building this model is to investigate the effects of various parameters (such as heavy goods vehicles proportion HGVs%, intergreen length and installation of red light cameras) on drivers’ compliance with the traffic light signal change and junction capacity as well as vehicles’ delays. Based on existing traffic simulation models such as CARSIM, the simulation methodology considered car-following algorithms with some modifications. These modified model includes the dilemma zone algorithms for predicting drivers’ STOP/GO decisions after the onset of amber. Various parameters were modelled such as distances from the stopline, travelling speeds, drivers’ responses to the signal change, junction width and the length of the amber period. The codes were written using FORTRAN-95 programming language. Traffic data from five sites were collected and analysed to be used for the calibration and validation of the developed model. The collected data included information about traffic flow characteristics, drivers’ compliance and junction details such as width, and the traffic lights periods and operation system (i.e. Fixed-Time (FT) or Vehicle-Actuated (VA) signals mode). Finally, the results of the newly developed model revealed that the number of signal violations increase as the intergreen length increases. Vehicle delays at junctions operated by FT signals are higher by 20% than those for VA signals mode. Moreover, an increase in the HGVs% causes a reduction in the red light running events by 40% and 45% at VA and FT traffic signal junctions, respectively. When the HGVs% constitutes of 50% of traffic composition, junction capacity is reduced by 42% and 51% at VA and FT junctions, respectively. In addition, the installation of red light cameras in the model showed positive effects on the reduction of signal violations. The reduction percentages were 70% at junctions controlled by VA signals and about 20% at FT traffic signal junctions.

Level dc:type.qualificationlevel

Doctoral (Level 8)

Year dc:date.issued

2018

Author dc:creator

• Al-Mukaram, NAR

Language dc:language

en

Identifier

oai:salford-repository.worktribe.com:1377947

OAI identifier oai:identifier

oai:salford-repository.worktribe.com:1377947