Complexity Science

1
Complexity Science Transport Systems
Jeffrey Johnson Joan Serras Design,
Development, Environment Materials The Open
University
to infinity and beyond
2
The Complex Systems Perspective Road systems -
interconnected systems of dynamic systems. Many
subsystems Populations, roads, railways, land
uses, regulations. All have multilevel
structure, e.g. people, families,

neighbourhoods, cities. All have dynamics gt
emergent dynamics of whole Subsystems have
thousands or millions of parts can only be
modelled on computers. How to represent and
modelling huge heterogeneous multilevel road
systems such as the whole of the UK or Europe ?
3
You came in here
You are here
or here
or here
4
You are here
5
(No Transcript)
6
(No Transcript)
7
(No Transcript)
8
(No Transcript)
9
Theres a lot of stuff
and youre still in England Wales !
10
Theres a lot of stuff
how can we represent so much stuff ?
11
Theres a lot of stuff
how can we represent so much stuff ?
12
Theres a lot of stuff
from micro to macro ?
13
Everything is becoming more more connected
Milton Keynes
Bedford
Traffic Modelling Conference
31-March-2009 Open University
14
Everything is becoming more more connected
Milton Keynes
Bedford
Traffic Modelling Conference
31-March-2009 Open University
15
Everything is becoming more more connected
Milton Keynes
Bedford
Traffic Modelling Conference
31-March-2009 Open University
16
Everything is becoming more more connected
Milton Keynes
Bedford
Cannot divide rule - the are connected !
Traffic Modelling Conference
31-March-2009 Open University
17

• Objectives of the research
• To model road system in an holistic way such
  that
• captures the dynamics of road traffic
• microdynamics e.g. shock waves
• macrodynamics e.g. transmission of congestion

Traffic Modelling Conference
31-March-2009 Open University
18

• Objectives of the research
• To model road system in an holistic way such
  that
• captures the dynamics of road traffic
• microdynamics e.g. shock waves
• macrodynamics e.g. transmission of congestion
• includes all the smallest level links nothing
  left out!
• aggregates bottom-up dynamics no information
  loss!

Traffic Modelling Conference
31-March-2009 Open University
19

• Objectives of the research
• To model road system in an holistic way such
  that
• captures the dynamics of road traffic
• microdynamics e.g. shock waves
• macrodynamics e.g. transmission of congestion
• includes all the smallest level links nothing
  left out!
• aggregates bottom-up dynamics no information
  loss!
• distributes computation and data naturally
  across
• administrative levels

Traffic Modelling Conference
31-March-2009 Open University
20

• Objectives of the research
• To model road system in an holistic way such
  that
• captures the dynamics of road traffic
• microdynamics e.g. shock waves
• macrodynamics e.g. transmission of congestion
• includes all the smallest level links nothing
  left out!
• aggregates bottom-up dynamics no information
  loss!
• distributes computation and data naturally
  across
• administrative levels
• illustrate by huge road systems, e.g. London,
  Europe

Traffic Modelling Conference
31-March-2009 Open University
21
Starting point it is like this
Traffic Modelling Conference
31-March-2009 Open University
22
Starting point it is like this
Plots of land zones
Traffic Modelling Conference
31-March-2009 Open University
23
Starting point it is like this
Activity Land Uses
Plots of land zones
Traffic Modelling Conference
31-March-2009 Open University
24
Starting point it is like this
Activity Land Uses
Plots of land zones
Traffic Modelling Conference
31-March-2009 Open University
25
destination
Starting point it is like this
origin
Trips are made between located activities
Traffic Modelling Conference
31-March-2009 Open University
26
destination
Starting point it is like this
origin
Traffic Modelling Conference
31-March-2009 Open University
27
destination
Starting point it is like this
origin
Traffic Modelling Conference
31-March-2009 Open University
28
destination
Starting point it is like this
origin
Traffic Modelling Conference
31-March-2009 Open University
29
destination
Starting point it is like this
origin
Traffic Modelling Conference
31-March-2009 Open University
30
destination
Starting point it is like this
origin
Traffic Modelling Conference
31-March-2009 Open University
31
destination
Starting point it is like this
origin
A route traverses a set of zones at different
levels
Traffic Modelling Conference
31-March-2009 Open University
32
Starting point it is like this
Plots aggregate into larger areas zones
Traffic Modelling Conference
31-March-2009 Open University
33
Zones aggregate into larger zones
Traffic Modelling Conference
31-March-2009 Open University
34
Zones aggregate into larger zones
Traffic Modelling Conference
31-March-2009 Open University
35
Zones aggregate into larger zones
Traffic Modelling Conference
31-March-2009 Open University
36
Larger zones aggregate into even larger zones
Traffic Modelling Conference
31-March-2009 Open University
37
Larger zones aggregate into even larger zones
Traffic Modelling Conference
31-March-2009 Open University
38
Larger zones aggregate into even larger zones
Traffic Modelling Conference
31-March-2009 Open University
39
Even larger zones aggregate into even larger zones
Traffic Modelling Conference
31-March-2009 Open University
40
Even larger zones aggregate into even larger zones
Traffic Modelling Conference
31-March-2009 Open University
41
and so on to zones for London, England, Europe,

Traffic Modelling Conference
31-March-2009 Open University
42
We assume there is a hierarchical set of zones
ZN-3 ZN-1 ? ZN ? ZN2 ? ZNk-1 ? ZNk
Traffic Modelling Conference
31-March-2009 Open University
43
We assume there is a hierarchical set of zones
ZN-3 ZN-1 ? ZN ? ZN2 ? ZNk-1 ? ZNk
kitchen
house plot, road segment
link between junction nodes
Traffic Modelling Conference
31-March-2009 Open University
44
We assume there is a hierarchical set of zones
ZN-3 ZN-1 ? ZN ? ZN2 ? ZNk-1 ? ZNk
kitchen
house plot, road segment
Abstract higher level lines (to be explained)
link between junction nodes
Traffic Modelling Conference
31-March-2009 Open University
45
Define a set of nodes - where a vehicle can cross
a road zone boundary
Traffic Modelling Conference
31-March-2009 Open University
46
Define a set of nodes - where a vehicle can cross
a road zone boundary
A link is defined by a pair of nodes
Traffic Modelling Conference
31-March-2009 Open University
47
Define a set of nodes - where a vehicle can cross
a road zone boundary
A link is defined by a pair of nodes conventional
links exist at Level N
Traffic Modelling Conference
31-March-2009 Open University
48
B
A
A Level-N route is a set of Level N links
Traffic Modelling Conference
31-March-2009 Open University
49
B
A
There are many N-Level routes between A and B
Traffic Modelling Conference
31-March-2009 Open University
50
B
A
There are many N-Level routes between A and B
Traffic Modelling Conference
31-March-2009 Open University
51
B
A
There are many N-Level routes between A and B
Traffic Modelling Conference
31-March-2009 Open University
52
B
A
Let this set of routes be a Level N2 Link
Traffic Modelling Conference
31-March-2009 Open University
53
A
B
20 routes
20 routes
20 routes
There are 8000 Level-N routes between A and B
Traffic Modelling Conference
31-March-2009 Open University
54
A
B
Put a N2-level nodes on the boundaries
Traffic Modelling Conference
31-March-2009 Open University
55
A
B
Make Level N2 links and a Level N2 route
Traffic Modelling Conference
31-March-2009 Open University
56
A
B
Make Level N2 links and a Level N2 route
Traffic Modelling Conference
31-March-2009 Open University
57
A
B
Make Level N2 links and a Level N2 route
Traffic Modelling Conference
31-March-2009 Open University
58
A
B
20 routes
20 routes
20 routes
We now have 20 20 20 1 61 ltlt 8000 routes !
Traffic Modelling Conference
31-March-2009 Open University
59
A
B
50 routes
50 routes
50 routes
We now have 100,000 possible routes !
Traffic Modelling Conference
31-March-2009 Open University
60
A
B
50 routes
50 routes
50 routes
We now have 150 9 159 ltltlt 100,000 routes!
Traffic Modelling Conference
31-March-2009 Open University
61
A
B
If the routes cross a Level N3 Zone ..
Traffic Modelling Conference
31-March-2009 Open University
62
A
B
make a Level N3 link as the set of N2 routes
Traffic Modelling Conference
31-March-2009 Open University
63
Hierarchical routes used for routing trips
Level N2 links
Level N1
Level N1 links
B
A
Level N links
Level N links
Within hierarchical zones, always use the highest
level links available
Traffic Modelling Conference
31-March-2009 Open University
64
Higher level links have lower variance !
Level N2 links
Level N1
Level N1 links
B
A
Level N links
Level N links
100
relative frequency
relative frequency
relative frequency
travel time, minutes
travel time
travel time, hours
Traffic Modelling Conference
31-March-2009 Open University
65
Hierarchical routes used for routing trips
Level N2 links
Level N1
Level N1 links
B
A
Level N links
Level N links
Note higher level links have more stable flow
and travel time statistics than lower level links
? initialisation heuristics ?
Traffic Modelling Conference
31-March-2009 Open University
66
TRANSIMS from Micro simulation to Macro dynamics
Traffic Modelling Conference
31-March-2009 Open University
67
TRANSIMS from Micro simulation to Macro dynamics
Traffic Modelling Conference
31-March-2009 Open University
68
TRANSIMS from Micro simulation to Macro dynamics
Traffic Modelling Conference
31-March-2009 Open University
69
TRANSIMS from Micro simulation to Macro dynamics
Traffic Modelling Conference
31-March-2009 Open University
70
TRANSIMS from Micro simulation to Macro dynamics
Traffic Modelling Conference
31-March-2009 Open University
71
Level N7 Level N6 Level N5 Level N4 Level
N3 Level N2 Level N1 Level N Level N-1 Level
N-2 Level N-3
The World ! Region (e.g. the Americas, Asia
) Nation (e.g. USA, UK ) State (e.g. Virginia,
Buckinghamshire) City, rural region (e.g. New
River Valley) Small town (e.g. Blacksburg, Milton
Keynes) Neighbourhood (e.g. VT Campus) Conventiona
l Links and zones Road segment, plot of
land House, garden, garage Rooms
Traffic Modelling Conference
31-March-2009 Open University
72

• Research in Progress
• Devise new microlevel simulation
• Implement at microlevel for Milton Keynes
• Implement hierarchical routing schemes
• New synthetic micropopulation all properties
• 5. Run the system for Milton Keynes
• 6. extend to UK
• 7. extend to Europe
• 8. extend to infinity, and beyond !

Traffic Modelling Conference
31-March-2009 Open University