Dynamic Traffic Assignment VIII: Applications of Simulation

1
Dynamic Traffic Assignment VIII Applications of
Simulation
Haris N. Koutsopoulos Northeastern
University M.I.T. Summer Professional Program
1.10s Modeling and Simulation for Dynamic
Transportation Management Systems  July/August
2003
2
Outline

• Introduction
• Microscopic Simulation Applications
• MITSIMLab
• Case Studies
• Mesoscopic Simulation Applications
• DynaMIT-P
• Case Studies
• Conclusion

3
Introduction

• Intelligent Transportation Systems (ITS)
• Characteristics
• Dynamic and Stochastic
• Complex Interactions
• Needs
• Decision Support Systems for operations (real
  time)
• System Design and Evaluation
• Emerging Tools
• Simulation-based

4
Simulation-based Evaluation

• Requires large scale integration of various
  components
• Ability to model at sufficient level of detail
• integrated networks
• traffic prediction
• travel behavior
• dynamic interactions
• Flexibility
• control strategies
• surveillance system
• sensitivity to design parameters
• Computational efficiency

5
Microscopic Traffic SimulationApplication Needs
6
MITSIMLab

• A simulation laboratory for ITS
• Evaluation
• dynamic performance (stability and robustness)
• interactions
• Design (ATMS, ATIS, APTS)
• Development of new technologies and algorithms

7
MITSIMLab A MIcroscopic Traffic SIMulation
Laboratory
- Traffic Management Center (TMS) -
Transit Operations Control Center (OOC)
- Traffic Surveillance Systems -
Transit Surveillance Monitoring Systems
(AVL, APC)
- Control and Routing Devices -
Transit Control and Traveler Information
Systems
MITSIM - Traffic Flow Simulator - Transit
Flow Simulator
8
Dynamic Traffic Management

• Traffic Control
• Intersection control
• Generic controller
• Transit signal priority
• Freeway control
• ramp metering
• mainline control
• Emergency response

• Route Guidance
• Variable Message Signs (VMS)
• Radio
• In-vehicle units

• Generic Overall Logic
• Pre-timed
• Reactive
• Predictive

9
Traffic Management Framework
10
Generic Controller
11
Bus Transit

• Supply
• Demand
• Transit surveillance monitoring
• Operations Control Center (OCC)

12
Bus Supply

• Supply
• Network
• Schedule and fleet assignment
• Driving behavior

Bus 1113
Route 43
Stop 33
13
Bus Demand

• Base level
• Stop-specific dwell time
• Aggregate level
• Average passenger
• arrival rate
• Alighting rate
• Disaggregate level
• O-D flow
• Passenger attributes
• Passenger choice

14
Bus OperationsControl Center (OCC)

• Schedule- and headway-based operations control
• Holding
• Transit signal priority
• Unconditional priority
• Conditional priority
• Schedule
• Headway
• Load

15
Travel Behavior

• Route Choice
• Pre-trip
• En-route
• Models
• path- based
• link-based

16
Incidents

• Location
• Start time
• Duration
• Severity
• Length
• Rubber-necking

17
Surveillance System

• Sensors and detectors
• point
• loop detectors, radar, over-height detectors
• point-to-point
• probe -vehicles
• area
• CCTV
• External Emergency Reports

18
Output and GUI

• Detailed data on
• speed
• flow
• queue length
• travel time
• vehicle trajectories
• sensor measurements
• Visualization/Animation
• Network editor

19
Details Matter Look-ahead

• Common approach myopic
• Drivers aware of next link only
• Problematic in urban settings
• Excess weaving and merging due to late lane
  changes

• MITISM path awareness
• Look-ahead
• Look-ahead distance distribution

20
Importance of Look-ahead

• without look-ahead with look-ahead

21
Importance of Look-ahead
Time Period

• MITSIM
• Without look-ahead
• With look-ahead
• Observed

Vehicle Count
Time Period
22
Computational Results

• Central Artery Network
• Experiment
• SGI Indigo2 R4400
• 143 nodes, 171 links
• 170 lane-kms
• 1 hour of operations

23
Applications

• Evaluation
• Traffic operations
• Intelligent Transportation Systems (ITS)
• Advanced Traffic Management Systems
• Advanced Traveler Information Systems
• Advanced Public Transportation Systems
• Emergency Response
• Design Refinement
• Algorithms
• Strategies
• Geometric Design

24
Framework
Evaluation loop
25
Discussion

• Scope of applications of micro-simulation
• Intersections ? networks
• Calibration
• Large number of parameters
• Data availability
• Interpretation of results
• Validation

26
Calibration
Demand
Supply

• Driving behavior
• Car-following
• Lane-changing and gap acceptance
• Desired speed distribution

• Dynamic OD

• Travel behavior
• Path choice set
• Habitual travel times
• Route choice model

27
Case Studies

• Applications
• Central Artery
• Stockholm
• Des Moines, IA
• Research and Development
• Evaluation of DTA systems

28
The Central Artery/Tunnel Network

• Approximately 110 lane-miles
• Loop detectors
• Lane Control Signals
• VMS

29
Emergency/Incident Management

• Incident in tunnel strategy
• Closure as soon as incident is detected
• Opening as soon as incident is cleared

30
No Delay
Delay in Opening
31
Lane Control Signs (LCS)

• Provides lane specific information to drivers
• Three states
• Green ?
• Yellow X
• Red X
• Reduction of delays due to weaving

32
LCS Design Options

• Design alternatives
• Design 1 2 Red X 2 Yellow X
• Design 2 1 Red X 4 yellow X
• Design 3 1 Red X 6 yellow X

• Various scenarios under incident conditions
• Conclusion
• 1 red, 4 yellow LCS configuration most effective

33
Application Example Evaluation of Ramp Metering
Algorithms

• An important control strategy to improve freeway
  flow

34
ALINEA

• Feedback control, local ramp metering algorithm
• Attempts to maintain a target occupancy on the
  mainline
• Simple, transferable, low implementation cost,
  efficient, flexible

35
FLOW

• Area wide control
• Metering rate function of
• Local Metering Rate (LMR)
• from Occupancy-Metering rate look-up table
• Bottleneck Metering Rate (BMR)
• based on influence zones (bottleneck sections)
• Queue length adjustments

Bottleneck Section
36
Case Study

• Very high OD demand

37
Results

• Effect of demand
• For ALINEA, travel time reduction at all demands
  except 80
• For FLOW, travel time reduction at 110 and 120
  demand
• Effect of bottleneck
• under no bottleneck, ALINEA is better.
• with bottleneck, FLOW is better (at very high
  demand)

38
Transit Signal Priority

• Stockholm
• 6 signalized intersections
• 4 bus stops
• headway 7.5 minutes

39
Validation Queue Lengths
40
Transit Signal Priority
41
Transit Signal Priority Results
100 Demand
140 Demand
42
Des Moines, Iowa

• Reconstruction of I-235
• Through downtown Des Moines, IA
• 16 Miles long
• Alternate Routes Parallel major arterials and
  Freeway
• Construction staging
• Network level impact of construction
• Congestion mitigation
• Signals, ramp metering and route diversion

43
Des Moines, Iowa

• LARGE Network
• 2500 Links, 200 Signalized and and 100
  un-signalized intersections
• 15, 000 O-D pairs
• 30, 000 vehicles at one time in the network

44
ATIS (DTA) Evaluation
45
Information
Sensor Data
MITSIMLab
46
Case Study Central Artery / Tunnel

• Impact of predictive information
• informed drivers
• O-D error
• Update frequency

47
Impact of Predictive Guidance
without guidance
with guidance
48
Results Informed Drivers

• Prediction Horizon 20 min

49
Results O-D Error

• Assumes 30 guided drivers

50
Results Update Frequency

• Assumes 30 guided drivers

51
DynaMIT-P

• Short-term planning
• Modeling elements
• Demand simulator
• Supply simulator
• Dynamic demand-supply interactions
• Characteristics
• Stochasticity
• Sensitivity to ATIS / ATMS designs
• Flexible time horizons

52
DynaMIT-P Applications

• Short-term planning applications
• Work zones
• ATIS/ATMS
• Evaluation of strategies
• Base case
• ATMS evaluation
• ATIS evaluation
• predictive
• instantaneous
• VMS
• Link (location) -based
• Path-based

53
VMS Evaluation

• Period 715 to 815
• Incident
• 720-745
• 60 capacity reduction
• Scenarios
• Base case (no VMS)
• Instantaneous VMS
• Predictive VMS

54
Average Travel Time (sec)
55
Impact of Incident

• No incident

Incident, no VMS
56
Impact of Incident

• No VMS

Predictive VMS
57
Conclusion

• Simulation is a valuable tool
• Increasing number of simulation applications
• appropriate scope and range
• Calibration
• aggregate
• transferability
• Education/Training
• Hardware-in-the-loop
• Operator-in-the-loop