these days I was asked about the impact of the DistanceTimeWeighting parameter (xServer2 request routing profile, used to be OPTIMIZATION in xServer1). A user played with different settings of this parameter while he kept every other setting stable (waypoints, profile, speed values,...).
I created the following images with a test client - maybe this info also enlightens you about the normal behaviour:
- Sample route from Karlsruhe (Southwest) to Heilbronn( East), 40t truck. DistanceTimeWeighting is iterated between 0 and 100 with steps of 10. With this scenario all routes above 40% follow the same track: use the highway A5 to Walldorf and then take the highway A6 to the east (90km, about 80mins)
The 4 other routes /distanceTimeWeight 0,10,20,30) follow a path through the countryside but there are some minor differences (more or less 70-80km, 90-100minutes).
- Now this screenshot uses a different base profile (car). With this profile the threshold value for the "highway" track (now 55minutes instead of 80mins) is now "above 50%". The country routes now take between 65 and 75 minutes (used to be between 90 and 100 minutes)
- more or less just for fun: a bicycle. The threshold is somewhere around 20%
- also applies to long distance by the way... be aware that many of the routes have a ridiculous distance (about 3km in total) which is caused by the ferries which do not contribute in distance (but in traveltime)
Best regards,
Bernd
