collision in sublane model with lcPushy > 0 (trac #3050)

[behrisch]

Migrated from http://sumo.dlr.de/ticket/3050

{
    "status": "closed", 
    "changetime": "2017-06-21T06:40:45Z", 
    "description": "", 
    "reporter": "namdre", 
    "cc": "", 
    "resolution": "fixed", 
    "_ts": "1498027245694473", 
    "component": "microsim - lane change", 
    "summary": "collision in sublane model with lcPushy > 0", 
    "priority": "major", 
    "keywords": "", 
    "time": "2017-05-03T14:48:20Z", 
    "milestone": "1.0.0", 
    "owner": "namdre", 
    "type": "defect"
}

[behrisch]

@namdre committed 0c8477f (aka r24692):
added test refs #3050

[behrisch]

@namdre committed 83544b9 (aka r24700):
making collision appear again. refs #3050

[behrisch]

@namdre commented:

collision is caused by numerical instability: MSVehicle::getLatOffset performs different computations depending on whether the vehicle is on the lane itself or only with it's back. This leads to inconsistencies in the occupied sublanes.

[behrisch]

• @namdre committed d1402cc (aka r24800):
  fix collision in sublane model with lcPushy > 0 (trac #3050) #3050
• @namdre changed owner from "" to "namdre"
• @namdre changed resolution from "" to "fixed"
• @namdre changed status from "new" to "closed"