Dear All
This is a copy of a reply that I sent to Alan re his query about bus
priority in Dubai. It has been suggested to me that I copy it to the
whole newsgroup so here goes ........
Dear Alan
We have undertaken a lot of work on active bus priority under
UTC/VA/SCOOT. The only consistent message is that the
benefits are highly site dependent. The TRL have a report on the
general subject, but I can't recall the reference. Maybe someone
else will provide you with this.
The principal means of assessing benefit the assessment for
priority on a linked signal network is by the use of TRANSYT, with
all sort of adaptations to assume what happens with and without a
bus (or other special Vehicle) during a particular cycle. As the
'special' vehicles do not arrive at the same time in a signal cycle, it
is very difficult to truly assess the benefit via a static tool.
TRANSYT does not assist the designer very much for assessing
where detectors should be placed, or what compensation timings
should happen after the bus, once a scheme is to be installed.
The presence of SCOOT will help in with the compensation
timings, but the assessment of the benefit of this is not easy.
Recently, we have been using microsimulation (mainly with
VISSIM, although other packages are available) to test benefits of
selective vehicle detection (SVD) and bus (and LRT) priority. This
has the major strength of modelling the actual (modelled) delays for
individual vehicles under differing conditions, using the actual
logical stage change conditioning of the signals. There is a lot of
work involved in this, but in the context of major capital
expenditure, it is well worth it. The microsimulation tool also
assists in the design process.
In Areas where the signal staging is fairly rigid (such as Dubai),
with each approach having a separate green, it would be
reasonable to expect that active bus priority would be pro rata
better than when 2 or more approaches have green together,
especially so if the green could be extended at the end of a stage
to ensure the bus does not have to wait most of the cycle for the
next green. Such a strategy of stage extension is likely to prove
better for other road users in perception terms as most drivers will
not realise that they have suffered a slight delay to accommodate a
few extra seconds for a special vehicle on another approach. As
the stage sequences is not unduly interrupted, delays to buses on
other approaches should be minimal. Implementation under
SCOOT should be fairly straightforward, provided you are able to
put in place suitable detection for the tagged vehicles (not always
easy on multi lane approaches).
Roger Hacker (a colleague at JMP) and I have submitted a
synopsis of paper to the IEE on the subject of SVD and
microsimulation. Hopefully it will be accepted for the conference in
March 2002.
A for good web sites, I have used the METRO one in West
Yorkshire a few times, and it would appear to be what you might
want. I see John Austin has offered some useful information, but
here's the METRO address (from memory)
http://www.metro-wyorks.co.uk/home.html
I hope this helps.
Regards and good luck
Adrian
JMP Consultants Ltd
Minerva House
East Parade
LEEDS
LS1 5PS
UK
Tel: 0113 244 4347
Fax: 0113 242 3753
Mobile : 07970 383195
The recipient of this email should ensure that the data
retrieved is correct and not corrupted and should be
compared against the hard copy which will have been issued
in parallel with this data to ensure its completeness.
JMP will accept no liability for any amendments
to the data contained therein, by any third party.
|