by bullock » Wed Nov 28, 2018 7:03 pm
Thank you.
At the bottom of the Southbound Details page of the TT&D report, there's a section labelled Cumulative Reports. In there, you can see that the CAS column of the Before-type runs includes a couple of outliers, including one run with CAS 19.2 mph, and another at 17.4 mph. These raise the average CAS to exceed that of the After-type runs.
The attached file assembles the detailed numbers and illustrates how they are averaged together.
For averaging the speed from link to link of a single run, Tru-Traffic uses CAS=CTL/CTT. But for averaging the speeds of separate runs along the same artery, Tru-Traffic calculates something like SUM(CAS)/COUNT(CAS), that is, it sums all the speeds together then divides by the number of runs. I believe this is the appropriate way to summarize multiple, separate runs.
If Tru-Traffic instead calculated the average CAS of separate runs by summing the CTL of all the runs and dividing by the sum of the CTT of all the runs (similar to the way it calculates the CAS or the average speed over multiple links along a single run), then it would indeed be surprising if the average CAS speed went down while the average CTT also decreased.
Regards.
Greg
- Attachments
-
Averaging CCT and CAS.pdf
- (35.62 KiB) Downloaded 1114 times