Quality Assurance
Some Examples
By evaluating the signal head
switching signals, the traffic engineer can visualize the traffic actuated
modification behavior of every node. There are some basic evaluation types
that will be shown in the next section.
Finally some examples will be
given of the application named "Green" used in Zurich.
Figure 1:
Cycle time
The figure shows the cycle time of
three signal heads. The left diagram shows the average cycle time
development over time using a 3 minutes interval. The right diagram shows
the distribution function of cycle time of the same three signal heads.
One can clearly see a strong mean
value and a distribution function around the mean value that is due to
public transport priority preemptions. The almost 0 cycle times helped the
technicians discover a measurement problem of the system!
The same evaluations can be done
for green times and for red times of the signal heads, as shown in the
next two figures.
Figure 2:
Green time
Figure 3:
Red time
The city of Zurich uses an online signal
head green time observation tool that is available through the
Internet. It uses Java applets on the browser side and a Java servlet on
the server side.
It is able to evaluate
-
mean green time of selected signal heads
-
statistical values per period as
- minimum green time
- maximum green time
- maximum red time
- variance of mean green time
- average waiting time for pedestrians
The evaluation intervals are defined as
-
last hour or a selected hour
-
last 24h or a selected day
-
last week of a selected week
The next screenshots show some
examples. At the moment the application can be called under http://www.verkehrsmanagement.ch/green/Green.jsp.
The application has not yet been published, so it will change its address
and access rights in the future.
Figure 4:
Measuring green time during one hour
8 Signal heads have been selected
for evaluations, each represented with another color. The upper part of
the panel contains a graphical diagram, the lower parts contains the
numeric values. At the moment signal head 576.S.17 is selected (red).
That's also why the red line in the diagram is thicker.
The statistical table values can
be copied and pasted e.g. into Excel for further analysis.
Figure 5:
Measuring green time during one day
The rose areas show invalid data.
The traffic lights have been switched off between 01:00 and 05:30. Yellow
shaded areas represent values with a lower level of confidence, e.g. due
to some missing measured values. There are no such intervals in the
figure.
Figure 6:
Measuring green time during one week
By evaluating the raw detector
signals, the traffic engineer gains a deeper insight into the traffic
behavior. Some basic evaluation types will be shown in the next section.
Finally some examples will be
given of the application named "Count" used in Zurich.
Figure 7:
Number of detected vehicles
The number of detected vehicles on
a list of detectors is a basic measure for determining traffic demand.
Figure 8:
Inter-vehicle distance
The vehicle count impulsions can
also be used for measuring the distance between
vehicles. The distribution function on the right diagram shows
clearly that most vehicles drive with a short distance to the next
vehicle. This means that there is mostly platoon
traffic detected by the shown detectors.
Figure 9:
Occupancy time
Occupancy time is used to estimate
the speed of the cars driving over the detectors. Higher speed means in
general shorter travel times that might be a political goal of the city
officials.
The city of Zurich uses an online traffic
count observation tool that is available through the Internet. It
uses Java applets on the browser side and a Java servlet on the server
side.
It is able to evaluate
-
count and occupation of one driving direction
-
counts of both driving directions
The evaluation intervals are
defined as
-
last 24h or a selected day
-
last week of a selected week
-
last month or a selected month
The next screenshots show some
examples. At the moment the application can be called under http://www.verkehrsmanagement.ch/count/Count.jsp.
The application has not yet been published, so it will change its address
and access rights in the future.
Figure
10: Measuring
count and occupancy during one day
The left figure shows count in red
and occupancy time in blue. The numeric values are shown on the lower part
of the diagram. They can be copied and pasted e.g. into Excel for further
analysis.
The right figure shows the count
of both directions (red and blue) of a street. The street has an
unbalanced load distribution: the flows do not compensate each other over
the duration of one day.
Figure 11:
Measuring count in both directions during one week and one month
The light blue shaded areas mark
missing measurement data.
There are a great number of
different evaluations that can be used for quality control. The chapter
shows some more examples.
-
The first example is a program used by the city of Zurich for
checking public transport priority behavior of the intersections.
-
The second example shows how the same evaluations could be done
without the technology installed in Zurich.
The city of Zurich uses an online public
transport regularity observation tool that is available through
the Internet. It uses Java applets on the browser side and a Java servlet
on the server side.
It is able to evaluate
-
time intervals between vehicles
-
travel time between stops
-
delays at stops and delays of vehicles in a synoptic diagram
The next screenshots show some
examples. At the moment the application can be called under http://www.verkehrsmanagement.ch/tram/TramJSP.jsp.
The application has not yet been published, so it will change its address
and access rights in the future.
Figure 12:
Measuring the time between two public transport vehicles
The left figure shows the time
between two vehicles following each other. The right figure shows the
distribution functions of these times, discretized to 1 minute. We see
mean 10 minutes distance between vehicles.
Figure 13:
Measuring travel time between two stops
The left figure shows how long a
vehicle usually drives from one stop to the next. The right figure shows
the distribution function of the travel time.
Figure 14:
Observing delays
This figure shows all trains of a
streetcar line, driving from bottom to top (the stop names are given at
the right, the departure time at the first stop is given at the bottom).
The diagram shows the delay of the anticipation of each vehicle at each
stop, using different colors (see the legend at the right edge of the
diagram).
-
In the case of horizontal contiguous areas there is a problem at
the stop.
-
In the case of vertical contiguous areas there is a problem with a
train.
By measuring delays of public
transport vehicles, the traffic engineer can discover priority treatment
problems at some intersections.
A GPS-bases trajectory analysis
system consists of the following parts:
-
The mobile unit which is installed in a public transport vehicle or
in any other probe car:
Figure 15:
Mobile GPS unit
The
mobile unit consists of a mobile computer (left) and a GPS receiver
(right).
-
The fixed unit with the evaluation program:
Figure 16:
Fixed GPS evaluation unit
The next figure shows two examples
of GPS trajectory evaluations:
-
Distance/Time (left): Delays can easily be seen in the graph when
identical trajectories are compared
-
Speed/Distance (right): The driven speed on the whole trajectory is
shown; several identical trajectories can be compared, but the distance
information is lost.
Figure 17:
GPS trajectory evaluations
More information can be found the the VS-GPS
product specification.
|
