[mkgmap-dev] avg speed on roads

Sat Apr 24 20:23:48 BST 2010


Florian Lohoff escribió:
> On Fri, Apr 23, 2010 at 06:31:42PM +0200, Carlos Dávila wrote:
>   
>> I think there are currently some tools in mkgmap to get a better time 
>> estimation. I have the lines below in my points style file, which 
>> certainly improve the arrival time estimation:
>> highway=traffic_signals { add mkgmap:road-speed = '-2'; add 
>> mkgmap:road-speed-min = '1' }
>>     
>
> I found the road speed stuff in the source already. It seems it is in the default
> style hardcoded by street type/classification which should be possible
> to override with some tag i think.
>   
It can be overridden with --ignore-maxspeeds, but you'll need to change 
lines style with something like this (for each type of road):
(highway=secondary | highway=secondary_link) & maxspeed>80    [0x04 
road_class=3 road_speed=5 resolution 20]
(highway=secondary | highway=secondary_link) & maxspeed>60    [0x04 
road_class=3 road_speed=4 resolution 20]
(highway=secondary | highway=secondary_link) & maxspeed>50    [0x04 
road_class=2 road_speed=3 resolution 20]
(highway=secondary | highway=secondary_link) & maxspeed<51    [0x04 
road_class=2 road_speed=2 resolution 20]
(highway=secondary | highway=secondary_link)            [0x04 
road_class=2 road_speed=4 resolution 20]
The kind of road you describe in La Palma usually have some curvy or 
urban sections with actual maxspeed lower than the generic one for the 
road. If you tag them you'll get better time estimation.
As an example of what I mean, you can have a look at [1] where I get 
quite accurate estimation on my nuvi.
[1] http://www.openstreetmap.org/browse/relation/12953
> And a maxspeed has nothing to do with what i hav a problem with. All the
> roads i travel dont have a maxspeed set, are primary and thus for the 
> time calculation my GPSMap 60 takes something like 80 km/h - which on 
> la palma is on some parts more than twice the actual speed one can 
> travel ... An its nothing time dependent - its something which depends
> on the physics of the street - tight corners reduce the speed beeing 
> traveled ...
>
> Typical speed/ average speed would be more accurate ...
>