jam front propagation to low (trac #2244)

[behrisch]

currently the jam front propagates so slow that the subsequent segment has emptied completely before the next vehicle may enter it. This contradicts the fact that the vehicle should enter when the predecessor has just started moving (and still has 85m to travel before leaving the segment).

Migrated from http://sumo.dlr.de/ticket/2244

{
    "status": "closed", 
    "changetime": "2016-05-04T14:03:03Z", 
    "description": "currently the jam front propagates so slow that the subsequent segment has emptied completely before the next vehicle may enter it. This contradicts the fact that the vehicle should enter when the predecessor has just started moving (and still has 85m to travel before leaving the segment).", 
    "reporter": "namdre", 
    "cc": "", 
    "resolution": "fixed", 
    "_ts": "1462370583881561", 
    "component": "simulation general (meso)", 
    "summary": "jam front propagation to low", 
    "priority": "major", 
    "keywords": "", 
    "time": "2016-04-14T13:17:16Z", 
    "milestone": "0.27.0", 
    "owner": "namdre", 
    "type": "defect"
}

[behrisch]

@namdre committed 824ab93 (aka r20448):
added test refs #2244

[behrisch]

@namdre commented:

Suggested solution is to introduce a new parameter (tau_s) for the back-propagation time gap.
Theoretical considerations suggest tau_s ~ tau_ff - 7.5/vMax

[behrisch]

@namdre commented:

Replying to [comment:2 namdre]:

Suggested solution is to introduce a new parameter (tau_s) for the back-propagation time gap.
Theoretical considerations suggest tau_s ~ tau_ff - 7.5/vMax

no new parameter is necessary as tau_jj already fills this role

[behrisch]

@namdre commented:

Replying to [comment:3 namdre]:

Replying to [comment:2 namdre]:

Suggested solution is to introduce a new parameter (tau_s) for the back-propagation time gap.
Theoretical considerations suggest tau_s ~ tau_ff - 7.5/vMax

no new parameter is necessary as tau_jj already fills this role

Except, the equation for the jam/jam case forces headyway to be n*tau_jf if n is the maximum number of vehicles for the segment, thus enforcing the contradictory behavior. (This exaclty follows the original model design (Eissfeldt dissertation, Equation 3.23)

[behrisch]

@namdre committed 622cc19 (aka r20469):
added failing test refs #2244

[behrisch]

@namdre committed e7b62e3 (aka r20470):
tweaking test refs #2244

[behrisch]

@palvarezlopez committed 6ba0eb1 (aka r20473):
Merged revision(s) 20332-20472 from trunk/sumo:

[behrisch]

• @namdre committed a9b9740 (aka r20649):
  fix jam front propagation to low (trac #2244) #2244
• @namdre changed owner from "" to "namdre"
• @namdre changed resolution from "" to "fixed"
• @namdre changed status from "new" to "closed"

[behrisch]

@namdre committed c783e0d (aka r20650):
patching expected results refs #21, refs #2244