Skip to content

Project Meeting 2019.03.22

Ben Stabler edited this page Mar 22, 2019 · 23 revisions

Verification

Summaries

  • See verification table for details
  • Value of time distribution is similar
  • Work county to county flows are similar
  • Auto ownership by taz is similar
  • CDAP results are similar except for preschool kids (see below)
  • Indiv mandatory tour freq is similar except for preschool kids
  • Indiv mandatory tour departure and duration match pretty well except for some differences in early versus late times for work
  • Joint tour freq matches pretty well, except we have less tours. This may be related to the preschool kids issue.
  • Joint tour composition matches pretty well, except we have less tours. This may be related to the preschool kids issue.
  • Joint tour distribution matches pretty well, except we have less tours and less short distance tours. This may be related to the preschool kids issue.
  • Non-mandatory tour freq does not match well (see below)
  • Tour mode choice matches reasonably ok, but there are too many tours because of the issue above

Issues / ToDos

  • TM1 does model mandatory travel for children too young for school. The documentation and size terms suggest otherwise, but there are mandatory tours for preschool kids. I plan to turn it back on to be consistent. We probably want to add an additional school segment - preschool - to make things clearer for the next developer/user. This is what TM2 does.
  • I reviewed all ASIM expressions against TM1 UECs and identified three differences:
    • parking costs need to be turned
    • all zones assumed to have short walk access to transit since we decided long ago to not implement transit subzones since multiple-zones is a superior approach. This will cause differences in the verification.
    • free parking is not implemented, including downstream expressions
  • @Ben currently tracing calculations for indiv non-mandatory tour freq since something isn't right
  • @Jeff to implement parking costs and free parking and other fixes as identified
  • @Ben to continue verifying / tracing

Project admin

  • Welcome Oregon DOT to the consortium!
  • Jeff added sampling and logsums for non-mandatory travel
  • Jeff finished implementing the atwork sbutour freq expressions
  • Cost to complete to add parking costs and free parking model is $5k (~ 20 hours of Jeff)

Shadow pricing by iteration scatter plots

Clone this wiki locally