Robustness and Reliability: uncertainties at calamities

1
Robustness and Reliability (un)certainties at
calamities

• The robust use of Dynamic Traffic Management at
  calamities

Travel Time Reliability
Victor Knoop
Transport Planning
2
Overview

• Recent research projects
• Spillback, blocking and capacity conclusions
• Future work

3
Recent research

• Simulation of evacuation of Rotterdam area
• Urban roads and motorways
• What is influence of spillback in combination
  with Dynamic Traffic Management
• Many models without spillback
• What is influence of capacity reductionin
  combination with Dynamic Traffic Management
• In evacuation capacity drop

4
Capacity reduction

• Direct, delayed flooding makes tunnel unusable

5
Capacity reduction

• Calamities causes different behaviour

6
Characteristics of the simulation

• 240.000 cars, 44 origins, 6 exit points
• DSMART, model developed at TU Delft
• Macroscopic
• Spillback-effects
• Possible DTM-influence
• Destination update
• Path update

7
DSMART

• Marcroscopic
• Aggregated variables flow, density, avg. speed
• Base element cell, homogeneous
• Demand Supply

8
Screenshots DSMART

• Spillback

9
No spillback the process

• Queue on link will not affect outflow of upstream
  link
• gt Supply independent of density downstream link
• gt Change supply function for last cell of link
• Demand Supply

10
No spillback the visualization

• Enormous densities

11
Route and destination choice

• At start destination and path based on free flow
• Based on free flow times to all exit points
• Destination choice
• logit-like distribution
• the longer the travel time, the less people will
  go there exp(-mT)
• Path
• Probit model
• Several estimations (error!) of travel time,
  choose shortest in time small preference for main
  roads

12
Rerouting and Redestination

• Every 15 minutes
• Path update
• Path update and destination update
• Not only destination update (based on what?)
• Path updates probit model
• Destination update logit-like model
• Both only part of travelers change route /
  destination

13
Spillback effects

• No effect DTM measures in a model without
  spillback effects

14
Spillback DTM effects
15
Conclusion

• Dynamic Traffic Management Measures can sort out
  effect, but spillback must be simulated

16
Road blocking

• Direct attack on road, earthquake

17
Road blocking

• Block one link
• Investigate performance of network with blocked
  link
• Two scenarios
• Rerouting
• No rerouting
• Normal weekday OD

18
Results
With rerouting gt
lt Without rerouting
19
Reliability capacity reduction
20
Conclusion

• No analysis of derivatives, simulations
• Delay in the same order of magnitude as capacity
  reduction
• Speed reduction proportional with capacity
  reduction?
• Factor 1??

21
Future research

• DTM at calamities
• Flooding, snow, earthquake, nuclear disaster
• Peoples reaction at a calamity
• Special DTM measures
• Which measures
• What are peoples responses to this measures
• Which reliability at calamities (no electricity,
  no gsm)
• Could one account for bad prediction in applying
  tactic
• Which performance indicator
• New (microscopic) simulator basis?

22
No spillback
Back
23
DSMART

• Marcroscopic
• Base element cell, homogeneous
• Demand Supply