SIMULATION OF PEDESTRIAN MOVEMENT IN SHOPPING STREET SEGMENTS

1
SIMULATION OF PEDESTRIAN MOVEMENT IN SHOPPING
STREET SEGMENTS

• Aloys Borgers, Inger Smeets,
• Astrid Kemperman, Harry Timmermans
• Urban Planning Group
• Eindhoven University of Technology
• The Netherlands

2
Models of pedestrian behavior

• Regression Analysis
• Space Syntax
• From entry to shop . to shop . to exit
• Microscopic models (Cellular Automata,
  Multi-Agent Systems, .)
• Etc

3
Space Syntax models
4
From origin to destination (Borgers
Timmermans, 80s)
2
3
Exit Start
1
5
From link to link (Borgers Timmermans, 00s)
6
Microscopic
7
Purpose Modeling pedestrian behavior in shopping
street segments
shops
shops
8
Shopping street segment
shops
Entry Exit
Entry Exit
shops
9
Network
10
Shopping street segment network
shops
Entry Exit
Entry Exit
shops
11
Representing routes
shops
shops
12
Modeling pedestrian behavior
Main principle from current link to adjacent link
13
Main attraction of exits
Entry
14
Secondary attraction of exits
15
Main attraction of shops
16
Secondary attraction of shops
17
Attraction of zones
18
Walking on the right side
19
Stop in shop
X
20
Modeling choice of link

• Pl exp(Vl) / ? l exp(Vl)
• Pl probability link l will be chosen from all
  adjacent links
• Vl utility of link l
• Vl ?k Xlk

21
X-variables

• Type of link
• In transfer zone
• In center zone
• walk on right-hand side

22
X-variables

• Attraction of exits (primary and secondary)
• Attraction of shops (primary and secondary)
• fashion
• shoes
• department stores
• fast food / drinks
• books
• electronics

23
X-variables

• Stop in shop
• branch specific constants
• branch specific floor space

24
Data collection

• One week workshop in July 2004
• Antwerps main shopping street
• Students
• Observed physical characteristics
• Counted pedestrians
• Tracked pedestrians

25
De Meir Antwerps Main shopping street
26
De Meir Antwerps Main shopping street
Segment B
Segment A
27
Segment A
Clothing Shoes Fast Food
28
Segment B
Clothing Shoes Dept Store Books Electronics
29
Tracking pedestrians
30
Estimation of parameters

• Number of tracked pedestrians
• Segment A 157
• Segment B 176
• Estimation Limdep
• Rho-square 0.73

31
Attraction of exits segment A
32
Attraction of exits segment B
33
Attraction of zones
0.0 -3.2 -0.4 -3.2 0.0
34
Attraction of walking on the right side
35
Attraction of shops
primary secondary
Clothing -
Shoes - -
Dept Store - -
Fast Food - - - - - - - - - -
Books
Electronics
36
Stop in shops
constant floor space
Clothing
Shoes
Dept Store
Fast Food
Books
Electronics
37
Simulation

• Start from observed starting position
• 50 times per route
• Mean absolute difference per link per simulation
  
• Segment A 3.20
• Segment B 4.06
• Mean route length Observed Simulated
• Segment A 110 m 105 m
• Segment B 90 m 79 m

38
Segment AObservedSimulated
39
Segment BObservedSimulated
40
Conclusions Recommendations

• Link to link approach
• Model performs pretty good
• Extend model
• Use complete shopping trips
• Include effect of having visited a shop on same
  type of shops
• Replicate
• Other street segments
• Other shopping areas
• Other conditions