MSc Thesis Proposal
Modeling and control of highway traffic
Keywords: traffic, modeling, control, optimization
Traffic jams are becoming more and more
acute every day.
They do not only
costs due to unproductive time losses;
they also augment the possibility of accidents
and have a negative impact
on the environment (air pollution, increased fuel consumption)
and on the quality of life (noise, stress).
Since traffic congestion
is such a pressing problem,
there certainly is a need
for measures that can be implemented
on the short term. One of such short term measures
is to augment the capacity of the existing
infrastructure by regulating
and redirecting the traffic flow.
In this proposal we
modeling and control of traffic flows on highways.
We shall start by studying several models that
describe the traffic flows on the highway
and then select one or more that best suit our aims. Next, we
shall investigate how control measures such as
information panels, variable message signs,
ramp metering (see figure below),
advisory speeds, etc. can be used optimally
to decrease the lengths and the frequency of occurrence
of traffic jams.
The operation of many contemporary traffic controllers is based on the
local traffic situation. E.g., ramp metering can be switched on when
the traffic around the on-ramp is becoming too dense, speed limits are
imposed when there is congestion downstream, or route information is
given based on congestion in the vicinity of the DRIPs (Dynamic Route
Information Panels). While these control measures are aimed at
solving or alleviating the local problems, they also have an effect
further away (and thus later) in the network. E.g., solving a local
congestion can create another congestion somewhere else in the
network. Therefore, considering the future network effects of these
control measures can considerably improve the flows on the network.
Model predictive control seems to be a good approach
to tackle this
problem, since predictions about the future behavior and the
development of traffic flows can be considered.
If you are interested in selecting this project as your MSc project,
please come along or send an email for more information.
This page is maintained by
Bart De Schutter.