Got another question....

I am getting two very extreme results from the Synchro queue and SimTraffic queue. Synchro gives me a 40 queue 95th %-tile queue (not "m" or "#" queue), but SimTraffic simulated a 400 feet 95th %-tile queue. What could cause that different?

Model is calibrated, simulated multiple times with different random numbers to obtain the average. Very low denied entry.

Thanks again.

Is it signaliuzed or unsiganlized. Synchro and simtraffic are two different models. it has been a very common misconception among traffic engineers that they would produce same results (as they are provided by same compnay). for many jurisdictions some of the traffic engineers they even interchangeably refer these two models. One of my traffic study was rejected because I did not run synchro for an hour (s/he meant simtraffc)
Is it signalized or unsignalized? If signalized is it corrdinated or uncoordinated? Is there a ramp up or down? have you watched the simulation? did you see any unusual behavior?
before deciding which one would be thrown out, several questions need to be answered.

quote:

Originally posted by Jahid Russel:
Is it signaliuzed or unsiganlized. Synchro and simtraffic are two different models. it has been a very common misconception among traffic engineers that they would produce same results (as they are provided by same compnay). for many jurisdictions some of the traffic engineers they even interchangeably refer these two models. One of my traffic study was rejected because I did not run synchro for an hour (s/he meant simtraffc)
Is it signalized or unsignalized? If signalized is it corrdinated or uncoordinated? Is there a ramp up or down? have you watched the simulation? did you see any unusual behavior?
before deciding which one would be thrown out, several questions need to be answered.

It is a coordinated signalized intersection in the middle of a one-way corridor. I had done all the suggestions, nothing unusual and had thrown out the bad results (and the too good). Only the average results are counted. And the model has calibrated to look very close to the actual field conditions.

I guess my question here is what is the methodology behind this 50th and max queue calculations in SimTraffic. I still can say the Synchro queue is based on the HCM, and can be powerful; but when saying SimTraffic queue is a computer simulated result; it's less convincing to the reviewer without explaining the methodology some time. May be this is something John can explain? I know it won't be something simple, but a rough idea is still better than nothing.

queue calcs can be tricky in simtraffic. queues extending over multiple bend nodes or intersections can cause misintrepretation of the data.

example:
We were modeling a new interchange. Existing overcrossing converted to a tight diamond. 2 New signals at intersections. primary queue was to the south of the highway, but about 3 driveway entrances (large apartment complexes)modeled as part of the simulation. Queue extended past all of these driveways for a time. how do you figure queue?

Since simtraffc figures queues based on the distance from the intersection, we got several different queue lengths. It looked pretty good if you look at the signal queue report, but that link was only a couple hundred feet long. watching the simulation revealed the max queue was over 1000 feet. I wound up printing a scale image of the simulation including background, and marking the end of the queue every signal cycle, and making the measurements manually.