 |
Work in collecting individual loop transitions was first experimented
with as part of the I-880 Field Experiment to assess the benefits
of Freeway Service Patrols. The Field Experiment collected peak
period data from 19 detector stations installed on I-880 and equipped
with dual loop detectors, extensive incident data, and probe vehicle
travel times for 50 week days. Rather than aggregating the detector
data into discrete samples, the database contains event data (i.e.,
individual detector actuations) sampled at 60 Hz [29,31,32]. Because
the database is so rich and the fact that all of the data are available
over the Internet, it has become an important resource for traffic
research around the world. Many lessons were learned during implementation
and subsequent analysis. The implementation itself was labor intensive,
the existing communications infrastructure could not transmit the
high resolution detector data, so these had to be collected manually.
The database was sufficient to assess the benefits of Freeway Service
Patrols, but it may not be detailed enough for other applications.
For example, with only 50 days, the data set may not be sufficient
for long term studies. Next, the data only includes peak periods
and does not have any weekend data. The headway between probe vehicles
may be too long to verify some applications. Finally, each probe
vehicle and detector station had a clock, but these clocks were
not coordinated accurately.
The standard Caltrans detector station uses a Model
170 controller sampling the loop sensors at 60 Hz. Under normal
operation, the 60 Hz event data are internal to the controller,
and the output data are typically aggregated into 20 second or 30
second average velocity, flow, and occupancy measurements. Caltrans
developed new controller software for the I-880 Field Experiment
that preserves the 60 Hz event data. Because the Model 170 controller
is based on 20 year old technology, simply outputting this data
stream consumes all of a controller's processing power. The I-880
Field Experiment used a laptop computer, in conjunction with each
controller, to collect and store this data stream in the field.
The BHL uses the same controller configuration, but rather than
storing the event data locally in the controller cabinets, it is
transmitted to the University of California via a wireless CDPD
modem and the Internet. The controller generates a new data packet
each second containing all the loop transitions seen by the controller
in the previous second. Finally, the I-880 Field Experiment was
constrained by limited storage capacity on the laptops; as a result,
data collection was restricted to weekday peak periods. In contrast,
the new laboratory uses a central server with enough disk space
to store over 3,000 hours of data from each station in the BHL.
The database is regularly backed up to CDROM, thus, allowing the
new system to monitor the freeway continuously, 24 hours a day,
seven days a week, indefinitely.
Additional enhancements to the BHL loop detector system were investigated
and implemented in 2002 under PATH Task Order 4134, “Loop
Detector Data Collection and Travel Time Measurement in the Berkeley
Highway Laboratory.” Several single and dual loop diagnostic
tests were developed, allowing real-time evaluation of the status
of loop detectors. Vehicle reidentification techniques were improved,
resulting in higher frequency of reidentification while simultaneously
reducing error rate. A less expensive, simpler, and more robust
communications system was implemented using off-the-shelf components.
Also, BHL loop detector system technologies were modified to make
them more adaptable to different cabinet, station, and loop configurations
so that a BHL-style system could be more easily deployed in new
locations.
|
