Hello Tom, the jumpy issue should be fixed in revision [19345]. It was related to sub-second simulation (http://sumo.dlr.de/trac.wsgi/ticket/1440). Thanks for supplying the scenario.
The connection issue is a caused by a known design problem in the heuristic for lane guessing which I hope to fix soon. Also, explicit connections are currently not imported from OSM ( http://sumo.dlr.de/trac.wsgi/ticket/1990) so you currently have no option for specifying them in your OSM input. The preferred work-around would be to either a) specify additional .con.xml files when doing the OSM import ( http://sumo.dlr.de/wiki/Networks/Building_Networks_from_own_XML-descriptions#Explicitly_setting_which_Edge_.2F_Lane_is_connected_to_which ) b) specify your network in sumo plain-xml files ( http://sumo.dlr.de/wiki/Networks/Building_Networks_from_own_XML-descriptions ) c) design it with NETEDIT. ( http://sumo.dlr.de/wiki/FAQ#How_can_I_obtain_NETEDIT.3F) regards, Jakob 2015-11-12 6:36 GMT+01:00 Lockhart, Thomas G (398I) < [email protected]>: > afaict the issue is still present. I updated using: > svn update > make -f Makefile.cvs > make > > Here is my test case, with a six-lane intersection generated from a > hand-built OSM file. The README has the commands I used to generate the > derived files and run the test case. Hopefully I got those about right. > > Note that the east-west road has only a single lane of through traffic, so > the jumping behavior is limited to the north-south road which has three > lanes of through traffic. (I would like to convince netconvert to have both > roads with three lanes of through traffic; suggestions appreciated ;) > > Look for a northbound vehicle jumping between the left two lanes at about > time 60. > > Thanks! > > - Tom > > ------------------------------------------------------------------------------ _______________________________________________ sumo-user mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/sumo-user
