Trip Log distance

Questions and discussions about operating Tru-Traffic

Moderator: bullock

Trip Log distance

Postby Chris » Wed Jul 31, 2013 2:40 pm

I am reviewing some travel logs results and I am starting to pay closer attention to the CTL column to make sure all my runs are within 1-3% of other runs. This kind of alerts me to possible noise in my runs or general problems with my model. On a particular model I have recorded an hours worth of runs and have let TruTraffic generate reports on the individual runs. As I look through the trip points I noticed that "previous node" and 'Next node" are not foloowing the pattern. For instance i expect the pair to go -1,12,23,34,45,55,54,43,32,21,11 but occassionally I am getting 12,23,34,45,54,43,32,21, I am not getting the details to show a portion where the previous node and next node are the same node, which I believe is normal.

here is the file
Attachments
S Fred email.Dgm
(125.74 KiB) Downloaded 636 times
Chris
 
Posts: 59
Joined: Wed Nov 17, 2004 2:25 pm
Location: San Antonio, TX

Re: Trip Log distance

Postby bullock » Thu Aug 01, 2013 5:13 pm

Well, that's interesting.

Tru-Traffic is getting confused by some noise in the GPS readings around the stop bar on the return trip (when the vehicle is about to reenter the artery for the next run). I'll give you a workaround for the meantime, but the next minor update to the software will smarten it up a bit to avert the problem in this case.

When the trip log seems to cross an intersection multiple consecutive times, the software has to decide whether

  1. the vehicle made a U-turn within the intersection (which should count as just one crossing), or
  2. the vehicle travelled downstream and did a (possibly midblock) U-turn, then reentered the artery (which should count as two crossings, marking the beginning of a new run in the opposite direction), or
  3. the vehicle passed through the intersection in one direction, then took a minor bypass route to reenter the artery in the opposite direction at the next intersection downstream (which should also count as just one crossing).

In your file, any idiot can see that this is case 2, so we should count two crossings. But Tru-Traffic isn't just any old idiot. While it has some logic to distinguish the above cases, it can be fooled. Some noise in the GPS readings around the stop bar, as the vehicle is about to reenter the artery, cause Tru-Traffic to (incorrectly) rule out case 2.

As I say, I'll smarten it up in the next minor update. In the meantime, here are a couple of workarounds:
  • Disable some of the points within the stop, just before the trip log reenters the artery. This may be simplest in the Trip Log Details Table. Scroll down the table, and uncheck the boxes for points #3370 - #3374, inclusive. That may be overkill, but it suffices to mask out enough of the GPS noise.
  • Split the trip log into two somewhere around the U-turn -- say, at point #3351.

Thank you for reporting the anomaly.

Greg
bullock
Site Admin
 
Posts: 218
Joined: Thu May 06, 2004 6:51 pm
Location: Pacific Grove, CA


Return to Program Operation

Who is online

Users browsing this forum: No registered users and 4 guests

cron