Microsimulation for Transport Planning Preliminary Findings

1
Microsimulation for Transport Planning-
Preliminary Findings

• YuChao Sun
• PhD Candidate
• The University of Western Australia
• PATREC Research Forum
• 19 September 2006

2
Contents

• Overview of microsimulation AIMSUN (video demo)
• Modelling building
• Model
• Testing
• Concluding remarks (video demo)

3
Overview of Traffic Microsimulation

• ABM (Agent Based Modelling)
• Discrete Time
• Stochastic (Give confidence interval rather than
  point estimate)
• Bottom-up

4
The Software

• AIMSUN NG (www.aimsun.com)
• User Friendly GUI
• objects colouring
• adjustable recordable animation
• tracking individual vehicles display floating
  data
• Fast network building editing
• unlimited number of links and vehicles
• Open Platform
• User-defined functions
• API
• SDK
• Computationally efficient
• Parallel computing with multiple threads
• Rich data outputs

5
Required Data

• Vehicle parameters
• Length
• Acceleration deceleration rate
• Desired speed
• Speed acceptance
• Give-way time
• Etc.
• Road parameters
• Capacity
• Speed limit
• Geometry
• Etc.

6
Required Data (Contd)

• Traffic control
• Fixed, actuated adaptive signal controls
• Warning signs
• Ramp metering
• VMS, etc.
• Public transport data
• Traffic demand (O/D matrix or turning
  proportions)
• Detectors

7
Transforming Macrosimulation Model

• Import shape files
• O/D matrices
• Traffic flow directions
• Lane numbers
• Connectors
• Curved Segments

8
Transforming Macrosimulation Model (Contd)
9
Transforming Macrosimulation Model (Contd)
10
Transforming Macrosimulation Model (Contd)
11
Transforming Macrosimulation Model (Contd)

• Junctions
• Control Plans
• Centroids
• Vehicle data

12
Transforming Macrosimulation Model (Contd)
13
Transforming Macrosimulation Model (Contd)
14
Transforming Macrosimulation Model (Contd)
15
The Model
16
Calibration

• Visual inspection
• Because of the unavailability of data, proper
  calibration could not be carried out
• Because of time constrains, insufficient number
  of replications have been run
• So the following tests are only used to
  demonstrate the possible applications

17
Section Performance
18
Section Performance
19
System Performance Measures

• Mean Flow Average number of vehicles passed
  through the network per hr
• Density Average number of vehicles per km for
  the whole network.
• Mean Speed Average speed for all vehicles that
  have left the system.
• Travel Time Average time a vehicle needs to
  travel 1km inside the network.
• Delay Time The average difference between the
  expected travel time (under ideal conditions)
  the actual travel time /veh/km.

20
System Performance Measures

• Stop Time average time at standstill /veh/km
• Number of Stops average number of stops /veh/km
• Total Travel total number of kms travelled by
  all vehicles left the network
• Total Travel Time total travel time experienced
  by all vehicles left the network.

21
Testing AlternativesVehicles with better or
worse info.
22
Testing AlternativesVehicles with better or
worse info.
23
Decision Table
24
Concluding Remarks

• Microsimulation models for large urban area such
  as Perth CBD are not only feasible but also
  beneficial (for both academics practitioners)
• One model can serve multiple purposes an
  integrated approach of planning management
  across different department
• There are more parameters in microsimulation
  models so a systematic approach is essential for
  calibration and validation.
• Transforming macrosimulation models can speed up
  the process, but the analysts should be aware of
  some methodological problems.

25
Concluding Remarks

• Although require more data, microsimulation
  models also produce more insights on how traffic
  evolves, e.g. time series data, sensitivity
  analysis, confidence interval estimate, etc.
• Although not calibrated, the model seems to
  suggest better information does not necessarily
  lead to better traffic conditions. This at least
  provides some words of caution to possible ITS
  applications.
• Thanks Questions?
• ?
• suny02_at_biz.uwa.edu.au