Developing a simulation model to evaluate the capacity of weaving sections

Al-Jameel, HAE

Abstract

Weaving sections have proved to cause bottlenecks on motorways because of the relatively
frequent lane changes in these sections. The use of analytical models in assessing weaving
section capacity (such as those suggested by the Highway Capacity Manual-HCM) has proved
to produce some inaccurate results. This could be due to the more complex nature of this
section compared with an isolated merging or diverging ones.
In this study, a new weaving micro-simulation model has been developed to focus on the
behaviour of drivers along a weaving section. This model consists mainly of four essential
algorithms to represent car-following, lane changing, gap acceptance and weaving rules.
The model was developed with data on drivers' characteristics at weaving sections which was
used in its calibration and validation processes. A large amount of data from loop detectors
obtained from the Highways Agency (i.e. the Motorway Incident Detection and Automated
Signalling-MIDAS) was collected. However, such data alone was found to be inadequate for
studying weaving section characteristics. Therefore, over 50 hours of video recordings were
collected from several weaving sections within the Greater Manchester Area. Such
comprehensive resource of field data helped in investigating other characterises such as
effective length, the percentage of segregation vehicles and merging/diverging points. The
results suggest that the effective length could be considered as about 200m for ramp weaving
sections where the actual weaving length is greater or equal to 300m. whereas this effective
length represents the whole weaving length when the actual weaving length is less than 300m.
In addition, the majority of drivers performed their merging as early as the first 100m from the
entrance point. Similarly, high proportions of diverging drivers entered the auxiliary lane very
close to the entrance point.
The newly developed model was applied in testing the effect on capacity of having different
percentages of heavy goods vehicles and a range of volume ratios. New management scenarios
were also applied using the developed model such as changing the configuration type of
weaving sections from Type A to Type B and shifting the point of diverging (i.e. moving to
the auxiliary lane) further downstream of the entrance point for a certain specified distance.
The results suggested an improved operational performance in terms of reducing the

Citation

Al-Jameel, H. Developing a simulation model to evaluate the capacity of weaving sections. (Thesis). Salford : University of Salford

• Thesis Type: Thesis
• Deposit Date: Oct 3, 2012
• Award Date: Jan 1, 2011