Mobility has always been a very important topic for society. With the growing amount of
traffic congestions, answers have to be found to solve them. Roughly 30% of all the traffic
congestions are so called shockwave traffic jams. These shockwave traffic jams are caused by
amplification of fluctuations in traffic.
To reduce these traffic congestions a lot of research focuses on so called Cooperative
Driving Systems (CDS). One example of this is: Cooperative Adaptive Cruise Control
(CACC). This is Adaptive Cruise Control (ACC) extended with communication. This
communication makes the vehicles able to react earlier on actions of their predecessors. Goal
of this system is to make a flow of traffic (or string of vehicles) so called ‘string stable’. This
is achieved when fluctuations in traffic are damped out in an upstream direction, what
prevents shockwave traffic jams. This improves overall throughput and therefore mobility.
There is over 20 years of literature available about ‘string stability’. However, almost
all of the literature is theoretical and assumes that there is a model available of the string
behaviour. In practice this is almost never the case, so it is difficult to apply this theory. This
research creates this missing link between practice and theory by proposing methods to
determine the vehicle following behaviour (and therefore the string stability). This is done
from measurement data obtained from practical tests.
This missing link, and the designed methods, are needed to be able to test a designed
vehicle following system like CACC. For example, after experiments are performed, to be
able to apply methods to validate the string behaviour and therefore the performance of the
system. This is a necessary step to validate, in the end, a finished product. Where after the
product can be implemented on the public road.
This research results in the design of two methods, one defined in the time domain and
one in the frequency domain. The time domain method is an analysis on measurement data
that uses the theory of norms: a norm analysis. The frequency domain approach uses the
theory of system identification to identify the string stability bode diagram from measurement
Both methods give positive results in their ability to determine the string stability from
measurement data. The system identification method results in more knowledge about the
model, where the norm analysis only results in knowledge about the analyzed piece of data.
On the other side, system identification requires some characteristics of the measurement
data, e.g. long data sequences and a frequency spectrum that contains a lot of frequencies.
This is necessary to determine an accurate bode diagram. The requirements for the
measurement data of the norm analysis are less strict.
This research does not only result in ways to determine string stability, but next to this,
more insight is gained in how to design proper tests. Proper tests, to test vehicle following
systems in a way that results in useful measurement data to determine the string behaviour.
Knowledge about the string behaviour is an indication of the performance of the system. The
link between practice and theory is crucial to implement CACC on the public road.
Traineeship report. - DC 2011.018