| Re: [sumo-user] How to make car following less safe? |
Hi Jan, the behavior you see is an artifact of conservative estimates in the sumo code, see the comments here: https://github.com/eclipse/sumo/blob/master/src/microsim/cfmodels/MSCFModel.cpp#L855 On the other hand there is https://github.com/eclipse/sumo/issues/4134 which seems to point in the opposite direction. You could try the DriverState https://sumo.dlr.de/docs/Driver_State.html to modify driver perception. This sdoes not apply to the accelerations so it should not be affected by this estimation. Best regards, Michael Am 15.10.20 um 02:15 schrieb grimm: > Hi there, a question on the car-following parameter apparentDecel: I've > been trying to change apparentDecel to values lower and higher than > decel in order to adjust car following towards more risky and safer > driving, respectively. While setting apparentDecel higher than decel > (e.g. 1.1 times decel for each vehicle type), I indeed see a decrease in > small time gaps, indicating safer driving. On the other hand, with lower > apparentDecel (ranging down to 0.7 times decel), there is no difference > (see chart below). Other parameters remain unchanged, mostly at their > default values (except for speed distribution and lengths). I'm using > the default car following and lange changing models with SUMO 1.6.0. My > simulation scenario is a motorway section. > > The following image shows the impact of the apparentDecel/decel ratio > (called sensitivity factor) on the mean individual braking time risk > (IBTR) in a motorway section, which increases with more frequent too-low > time gaps. > It appears that the car following model prevents this kind of unsafe > driving, using decel of the preceeding vehicle instead of apparentDecel, > although the documentation on > https://sumo.dlr.de/docs/Simulation/Safety.html states otherwise. Is > that possible? > > My goal is not necessarily to cause accidents, but to decrease time gaps > to a slightly unsafe level. Thus, I wouldn't disable the safety checks > in car following. The only other way to provoke more risky car following > would be to decrease tau slightly below action step length. However, > this quickly leads to large numbers of emergency braking maneuvers. > > Is there any other way to get higher numbers of risky time gaps I > haven't thought of yet? (Hints on literature sources also appreciated) > > Thank you very much! > Kind regards > Jan > ------------------------------------------------------------------------ > Sent from the Sumo User Mailing List mailing list archive > <http://sumo-user-mailing-list.90755.n8.nabble.com/> at Nabble.com. > > _______________________________________________ > sumo-user mailing list > sumo-user@xxxxxxxxxxx > To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user >
Attachment:
signature.asc
Description: OpenPGP digital signature