IMHO, the hardest part is not the recalculation of the optimal route. The
algorithms available are quite fast and reasonable. The most difficult part
is feeding accurate and up to date information into the model as a basis for
the calculations. There are some projects in the USA (and elsewhere, no
doubt) that provide real-time traffic information to travellers from the
Freeway Traffic Management Systems. That information has to be upgraded from
simple coloured displays on a computer screen indicating areas of slow and
fast traffic movement, to more detailed estimates of travel times for each
road link which could actually be used by a model to judge the fastest
route. Most of these systems only cover the freeways and not the arterial
road networks that would be used as diversion routes by motorists.
There may also be some applicability of artificial intelligence to this
problem. If you travel a short distance, update the information, and
calculate a new routing that is optimal from your current location to your
destination, the net result of these multiple decisions along the trip may
not work out to be the optimal routing that would have been chosen with the
benefit of hindsight and perfect information. When the radio station
announces that there has been a collision on the highway, the seasoned
commuter might think to himself "I could divert to XX Road because it's
empty now, but by the time I get there it will have filled up with everybody
else diverting off the highway so I'm better off just staying here (or
taking a third different route) instead".
--------------
Colin R. Leech
All opinions personal
[log in to unmask]
http://www.octranspo.com
> -----Original Message-----
> From: Husdal, J. [SMTP:[log in to unmask]]
> Sent: Sunday, March 05, 2000 7:44 AM
> To: [log in to unmask]
> Subject: real time route planning
>
> I am student at the MSc in GIS course at University of Leicester, UK, and
> I
> am in the process of narrowing down my topic for my dissertation, which
> will
> most likely be on Intelligent Transport Systems, Vehicle Location and
> Navigation, and Route Planning in particular.
>
> There is a lot of routing software available, but what I find is that they
> do not take into account the time of day when you are travelling, and that
> the traffic changes as you go along, so that the pre-planned route may not
> be the best route in the end.
> I would like to focus on real time route planners that work with dynamic
> networks and how this can be implemented. I know how to model a static
> network, but how do you model a network that is constantly changing, and
> how
> do you get the routing algorithm to recognise the changes and recalculate
> the route?
>
> If anybody is doing similar research, or knows of any leads or good
> starting
> points, literature or journal articles, please let me know.
>
> Thanks.
>
> Jan
>
>
> http://www.husdal.com - Geomatics and beyond
> Consulting and Applications Developing
>
> Jan Husdal
> A 403, Putney Road Houses, 33 Putney Road
> Leicester LE2 7TG, UK
> +44 116 215 2427 home
> +47 970 94 823 mobile
> [log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|