MSc Thesis Proposal
New ramp metering strategy for solving freeway traffic jams
Mentors:
A.
Hegyi and B. De
Schutter
Problem statement
Freeway traffic jams are often caused by
excessive traffic from the on-ramps. To limit the amount of traffic
entering the freeway often a traffic light is placed at the on-ramp,
the so-called ramp metering (see the figure for an example). This
traffic light periodically switches between red and green, and allows
one vehicle per green to the freeway. By regulating the switching
frequency the amount of traffic that enters the freeway can be
controlled. The resulting improvement of traffic flow observed in
practice varies between 0 and 10%, which is quite small.
Probably one of the reasons for this small improvement is that most
ramp metering algorithms do not take explicitly into account the
so-called capacity drop. The capacity drop means that the traffic flow
leaving a traffic jam is significantly lower than the flow under
normal free-flow conditions. The main consequence of the capacity drop
is that the traffic flow cannot be restored to the higher pre-jam flow
as long as the traffic jam exists. Therefore, the ramp metering
algorithm should explicitly aim at removing the jam. However, existing
algorithms do not explicitly take into account the existence of the
jam, but mainly aim at the prevention of traffic jams.
Goals
The goals of this MSc project are:
- Investigation of real traffic measurements of traffic jams at
on-ramps. Description of the process that causes the capacity
drop.
- Evaluation of existing ramp metering algorithms in the light of the
capacity drop phenomenon.
- Development of a new ramp metering
approach that explicitly takes into account the capacity drop and the
presence or absence of a jam near the on-ramp. Comparison of the new
approach with the existing ones.
Illustration of ramp metering and capacity drop.
If you are interested in selecting this project as your MSc project,
please come along or send us an email for more information.
This page is maintained by
Bart De Schutter.