May 8, 2011 (2:45PM~4:30PM) 13 th TRB National Transportation Planning Applications Conference

Integrated Dynamic Traffic and Transit Assignment Considering Traffic Congestion and

a Schedule-Based Transit Network 

May 8, 2011 (2:45PM~4:30PM)13th TRB National Transportation Planning Applications Confer-

ence

Hyunsoo Noh, Alireza Khani, Mark Hickman, and Yi-Chang Chiu 

The University of Arizona 

Contents

• Background and Motivation

• Dynamic Traffic Assignment Models

• Transit Assignment Model

• Integration Models

• Application (Rancho Cordova, CA)

• Result

• Conclusion

Background and Motivation

• Activity-based model

• Tour-based model 

• Sequential activity choice model 

• Necessity of time-dependent traveler model

• Necessity of intermodal architecture

Background DTA Model FAST-TrIPs Integration Model Application Result Conclusion

Dynamic Traffic Assignment Model

• DynusT and MALTA*

Background DTA Model FAST-TrIPs Integration Model Application Result Conclusion

AMS TDSP

DynusTMALTA

Relative-Gap Epoch

Assignment

GUI and Data Analysis

HTDSP

Relative-Gap Assignment

GUI

Transportation Operational PlanningReal-Time Traffic Operations

Postgres

DemandDemand

Network Partitioning and

Load Balance

Evac/Inci

Value Pricing

ITS

Intermodal

Work Zone

Ramp Metering

VISSIM Integration

* MALTA: Multi-resolution Assignment and Loading of Transportation Ac-tivities

Flexible Assignment and Simulation Tool for Transit and Intermodal Passengers (FAST-TrIPs)

• FAST-TrIPs• Frequency-based model• Schedule-based model

• Frequency-based Assignment• “Strategies” assignment approach• Headways for 10 minutes or less, or without the schedule

• Schedule-based Assignment• Route (specifically run) for transit schedule• Preferred arrival time (PAT) • Logit-based SUE model 

Background DTA Model FAST-TrIPs Integration Model Application Result Conclusion

Schedule-based Assignment Model

• Link-Based Time-Expanded (LBTE) Schedule Network

A B C Des

e1

e2 e7

e8

e10

e12

et

[25, 7]

Route 1Route 2

Route 3Route 4

[7, 1] [6, 28]

[4, 11]

[4, 15]

[10, 13]

[cost, departure time]

e3 [8, 5]

e4 [7, 10] e9[5, 16]

e6[6, 19]e5 [6, 8]

e14[10, 43]

e11[4, 21]

[10, 23]e13

Background DTA Model FAST-TrIPs Integration Model Application Result Conclusion

• Transit Network• Link-based schedule network• Google's General Transit Feed Specification (GTFS)

• Transportation Network• Access link

GTFS Stop Time

Trip IDArrival Time

Departure TimeStop ID

Stop SequenceTrip Headsign

…

GTFS Stop

Stop LatitudeStop Longitude

Stop CodeStop ID

…

Transportation Node

Node IDNode Latitude

Node LongitudeNode Type

…

Link-based Schedule

Link IDNode INode J

Link Type*Link Mode

Dep/Arr Time…

Background DTA Model FAST-TrIPs Integration Model Application Result Conclusion

* OD, Access, and Transit links

Integration Model 1

• Tour-Based Integration Tour-Based Model(Data: Trip, Tour)

traveler.dat(Auto, Transit,

and Auto-Transit)

DynusT

Auto Assignment

Transit (Intermodal)Assignment

Background DTA Model FAST-TrIPs Integration Model Application Result Conclusion

Intermodal Path

FAST-TrIPs

Auto – Transit

Simulation

PassengerSimulation

Auto, Transit Skim

Integration Model 2 (session 6)

• Activity Choice Model

Background DTA Model FAST-TrIPs Integration Model Application Result Conclusion

Activity-Based Model(O/D, Mode at Time Intv.)

Auto Assignment Transit (or Intermodal) Assignment

MALTAAuto + Transit Vehicle

Simulation

Re-Assignmentfor Missing

Arrival(Auto,Passenger)

Auto SkimTransit

(or Intermodal)Skim

Day-to-Day Day-to-Day

Convergence(Auto Assignment)

Convergence(Transit Assignment)

Auto Path(Intermodal)

FAST-TrIPs

FAST-TrIPs

MALTA

MALTAPassenger Simulation

FAST-TrIPs

Simulation Structure

Simulation of Vehicle

Movements(Auto, Transit) Passenger Simulation

Passenger arrival time, stop, boarding behavior

Passenger experience

Passenger arrival from auto

Background DTA Model FAST-TrIPs Integration Model Application Result Conclusion

DynusT | MALTA

Transit Passenger Assignment

Activities and travel requestse,g,, OpenAMOS, DaySim

Google GTFS and/or transit line information

Transit and intermodal trips Routes, stops, schedulesAuto trips

Dwell time, Hold

Transit vehiclearrival

Transit Passenger Skims and Statistics

FAST-TrIPs

Bus # 2321Pass. # 12Pass. # 25Pass. # 47

STOP STOPBus # 2322Pass. # 87Pass. # 19

…

Stop # S21Pass. # 24Pass. # 99

…

Stop # S33Pass. # 76Pass. # 59Pass. # 41

Transit vehicleapproach

Need to stop

Auto Skims

Auto part of Intermodal

Missing-a-bus New Path

Vehicle Skims

Application

• Auto Network (Rancho Cordova, CA)

Background DTA Model FAST-TrIPs Integration Model Application Result Conclusion

Road Network

Number of NodesNumber of Links

Number of ZonesNumber of Parcels

Number of Park-&-Ride

448 850 12120,486 3 (LRT)

Folsom Blvd & Mather Field Rd(cap.: 235)

Butterfield Way & Folsom Blvd(cap.: 406)

Trade Center Dr & Sunrise Blvd LRT(cap.: 487)

http://www.sacregion511.org/rideshare/parkandride/

• Transit Network (Rancho Cordova, CA)

Background DTA Model FAST-TrIPs Integration Model Application Result Conclusion

21

28

72

74

75

(LRT Gold line)

Transit Network

Number of StopsNumber of Trips

Number of Stop TimesNumber of Routes

163 1,20616,1175 Bus + 1 LRT

21

28

507

74

75

72

21 72 74 75 507

74 507

Folsom Blvd & Mather Field Rd

Trade Center Dr & Sunrise Blvd LRT

Butterfield Way & Folsom Blvd

Park-&-Ride Facility

Result

• Scheduled Hyperpath

Mode: Walk-Transit-Walk

75

507

Parcel # 499638PAT: 7:45 ~ 8:00

Parcel # 562770Folsom Blvd & Sunrise Blvd

On: 336347,7:28:00,7066Off: 336347,7:34:00,7069

Folsom Blvd & Mather Field Rd

On: 335305,7:40:00,370Off: 335305,7:48:00,9923/5290

On: 344131,7:44:00,370Off: 344131,7:52:00,5290/9923

9923 | FEMOYER ST & SCHIRRA AVE

MATHER BLVD & BLECKLEY ST | 5290

Access time : 2.92 Mins.

Transfer time : 0.85 Mins.

Background DTA Model FAST-TrIPs Integration Model Application Result f Conclusion

• Scheduled Intermodal Hyperpath

Mode: Auto-Transit-Walk

Background DTA Model FAST-TrIPs Integration Model Application Result Conclusion

9923 | FEMOYER ST & SCHIRRA AVE

MATHER BLVD & BLECKLEY ST | 5290

On: 335305,7:40:00,370Off: 335305,7:48:00,5290/9923

On: 344131,7:44:00,370Off: 344131,7:52:00,5290/9923

75

Folsom Blvd & Mather Field Rd

Access time : 4.86 Minutes.

Parcel # 499638PAT: 7:45 ~ 8:00

Parcel # 562770

• Simulated Passenger Paths

Mode: Walk-Transit-Walk

Background DTA Model FAST-TrIPs Integration Model Application Result Conclusion

Background traffic: 72,501 VehiclesTime: 6~10AM

Parcel # 499638PAT: 7:45 ~ 8:00

Parcel # 562770

Planned path for PassID 102 :444 (7:24AM) - walk - LRT Gold – Transfer at stop 7069 to Bus Route 75 - Alight at 5290

At time 447 (7:27AM)PassID 102 transferred to 7066 and got on 345158

MATHER BLVD & BLECKLEY ST | 5290

At time 453 (7:33AM) vehID 345158 arrived at 7069 passID 102 got off 345158 At time 454 (7:34AM) PassID 102 transferred to 370At time 459 (7:39 AM) PassID 102 got on 335305At time 467 (7:47 AM) vehID 335305 arrived at 5290 PassID 102 got off Arrived at destination

75

• Simulated Intermodal Hyperpath

Mode: Auto-Transit-Walk

Background DTA Model FAST-TrIPs Integration Model Application Result Conclusion

Parcel # 499638PAT: 7:45 ~ 8:00

Parcel # 562770

Background traffic: 72,501 VehiclesTime: 6~10AM

MATHER BLVD & BLECKLEY ST | 5290

75

At time 458 (7:38AM) PassID 102 transferred to 370At time 459 (7:39 AM) PassID 102 got on 335305At time 467 (7:47 AM) vehID 335305 arrived at 5290 PassID 102 got off Arrived at destination

Planned path for PassID 102 :454 (7:34AM) - Auto – Transfer at stop 7069 to Bus Route 75 - Alight at 5290

Conclusion

• More realized and integrated intermodal transportation among auto, transit vehicles and passengers

• Integration between activity model and DTA model

• On-going Works– “Missing-a-bus” – Convergence between different models– Full integration of transit, auto, and passenger assign-ment with activity models

Background DTA Model FAST-TrIPs Integration Model Application Result Conclusion

Acknowledgement

These projects are funded by • FHWA Exploratory Advanced Research (EAR) Program • SHRP2 C-10BOur collaborators are

FHWA EAR Program SHRP2 C-10B

Thank you !