Development of traffic micro-simulation model for motorway merges with ramp metering

Al-Obaedi, J

Abstract

This thesis focuses on the development of a micro-simulation model for motorway merge
sections. The aim is to study the effectiveness of applying some traffic management
controls and particularly focuses on applying ramp metering (RM) systems.
The new model has been developed based on car-following, lane changing and gap
acceptance rules. The model considered the multi-decisions undertaken by merging traffic
when a driver, for example, accepts the lead gap and rejects the lag gap. The cooperative
nature of drivers is also considered where motorway drivers allow others to merge in front
of them either by decelerating or shifting to other lanes (yielding) in the vicinity of
motorway merge sections. Video recordings, as well as data from the Motorway Incident
Detection and Automatic Signalling (MIDAS) were obtained from a selection of sites. The
data was used in the verification, calibration and validation processes of the developed
model. Other main sources of information include more than 4 million cases of successive
vehicles taken from UK motorway sites. These cases were analysed to study the effect of
vehicle types on the following behaviour for drivers. The main finding is that there is no
evidence that the average spacing between successive vehicles is significantly affected by
the type of leading vehicle.
Different RM algorithms have been integrated within the developed model. The results of
testing the effectiveness of RM controls using the developed model reveal the benefits of
RM in reducing time spent by motorway traffic (TTSM) but it significantly increases the
time spent by the merging traffic (TTSM). The overall benefits of implementing RM in
reducing total time spent (TTS) is limited to situations where the sum of motorway and
merge flows exceeds the capacity of the downstream section. Other issues related to RM
design and effectiveness have been tested such as the effects of having different durations
for peak periods, finding the optimum parameters for each algorithm, the effect of ramp
length (storage area) and the effect of RM signals position. The results suggest that RM is
very efficient when implemented for short peak periods (e.g. less than 30 minutes). The
effectiveness of RM in decreasing the travel time for motorway traffic is increased with an
increasing ramp length but with a significant increase in ramp traffic delay. No significant
effect is obtained from altering the ramp signals' position.
Other tests include the use of other types of traffic management controls (e.g. applying
different speed limits and lane changing restrictions (LCR) at the approach to merge
sections). No significant improvements were obtained from testing different speed limit
values. The results suggest that LCR could reduce travel time for motorway traffic.
However, there are other practical considerations which need to be addressed before this
could be recommended.

Citation

Al-Obaedi, J. Development of traffic micro-simulation model for motorway merges with ramp metering. (Thesis). Salford : University of Salford

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