Presentation is loading. Please wait.

Presentation is loading. Please wait.

MOVING THE WORLD. Traffic Science Faculty "Friedrich List" Institute for Economics and Traffic Chair for Traffic Modelling and Econometrics.

Published by Modified over 9 years ago

Similar presentations

Presentation on theme: "MOVING THE WORLD. Traffic Science Faculty "Friedrich List" Institute for Economics and Traffic Chair for Traffic Modelling and Econometrics."— Presentation transcript:

1 http://www.helbing.org MOVING THE WORLD. Traffic Science Faculty "Friedrich List" Institute for Economics and Traffic Chair for Traffic Modelling and Econometrics Dresden University of Technology, Germany Stefan Lämmer traffic@stefanlaemmer.de Scaling Laws in the Spatial Structure of Urban Road Networks Physica A 363(1) 89-95, 2006

2 2 Institute for Economics and Traffic Chair for Traffic Modeling and Econometrics Stefan Lämmer traffic@stefanlaemmer.de 2 Scope of Study ► Urban road network analysis of 20 largest cities of Germany (ranked by population) ► Geographical database: Tele Atlas MultiNet TM City of Dresden with 9643 nodes and 22.307 links S. Lämmer, B. Gehlsen, and D. Helbing (2006) Scaling laws in the spatial structure of urban road networks, Physica A 363(1) 89-95.

3 3 Institute for Economics and Traffic Chair for Traffic Modeling and Econometrics Stefan Lämmer traffic@stefanlaemmer.de 3 Distinct Classes of Networks ► Random networks (Erdős and Rényi 1959) Exponential node-degree distribution High vulnerability to random failures ► Scale-free networks (Barabasi and Albert 1999) Short distances (small world phenomenon) High clustering coefficients Power-law node-degree distribution ► Urban road networks (Gastner and Newman 2004) Mainly planar cellular structure Limited node-degrees (average strictly less than 6) Very high network diameter, high redundancy

4 4 Institute for Economics and Traffic Chair for Traffic Modeling and Econometrics Stefan Lämmer traffic@stefanlaemmer.de 4 Generating a Road Network ► Combinatorial Optimization Problem Given: node positions and total link length. Minimize the average length of all shortest paths. Distances are the most fundamental property of road networks! (a) Link cost proportional to network distance (b)Link cost proportional to Euclidian distance Scale-free network Road network

5 5 Institute for Economics and Traffic Chair for Traffic Modeling and Econometrics Stefan Lämmer traffic@stefanlaemmer.de 5 Scaling of Neighborhood Sizes With a travel-distance-budget of r kilometers, a car driver can reach a neighborhood of size Dresden: d = 2.2 With a travel-time-budget of τ minutes, a car driver can reach a neighborhood of size

6 6 Institute for Economics and Traffic Chair for Traffic Modeling and Econometrics Stefan Lämmer traffic@stefanlaemmer.de 6 Effective Dimension ► Computation of effective dimension d: ► High values of d imply: Neighborhoods grow fast by small increases of travel-time budget. High accessibility of distant places. Heterogeneous distribution of road speeds, e.g. the city might have an underlying highway system. Typical values for d Berlin2,330Bremen2,220 Hamburg2,350Duisburg2,050 Munich2,463Leipzig2,304 Cologne2,372Nuremberg2,399 Frankfurt2,388Dresden2,205 Dortmund2,091Bochum2,279 Stuttgart2,008Wuppertal2,040 Essen2,243Bielefeld2,337 Düsseldorf2,700Bonn2,134

7 7 Institute for Economics and Traffic Chair for Traffic Modeling and Econometrics Stefan Lämmer traffic@stefanlaemmer.de 7 Spatial concentration of traffic flow ► Traffic flows can be estimated by betweenness centrality b, which is the number of shortest paths visiting a link or a node. (We assume homogeneous OD-flows and ignore congestion effects)

8 8 Institute for Economics and Traffic Chair for Traffic Modeling and Econometrics Stefan Lämmer traffic@stefanlaemmer.de 8 Scaling of traffic flow on nodes ► High values of β imply: Traffic flow concentrates on few highly important intersections. Low redundancy (lack of alternative routes) High vulnerability to failures of traffic control Typical values for β Berlin1.481 Hamburg1.469 Munich1.486 Cologne1.384 Frankfurt1.406 Dortmund1.340 Stuttgart1.377 Essen1.368 Düsseldorf1.380 Bremen1.351 Duisburg1.480 Leipzig1.320 Nuremberg1.420 Dresden1.355 Bochum1.337 Wuppertal1.279 Bielefeld1.337 Bonn1.374

9 9 Institute for Economics and Traffic Chair for Traffic Modeling and Econometrics Stefan Lämmer traffic@stefanlaemmer.de 9 Concentration of Traffic on Road Meters ► Lorenz-Curve ► High values of Gini-Coefficient g imply Bundling of traffic on arterial roads Existence of bottlenecks (bridges) Reduced traffic (in residential areas) Distinct hierarchy of roads Typical values for g Berlin0,871Stuttgart0,894Nuremberg0,854 Hamburg0,869Essen0,892Dresden0,870 Munich0,869Düsseldorf0,849Bochum0,847 Cologne0,875Bremen0,909Wuppertal0,881 Frankfurt0,873Duisburg0,900Bielefeld0,872 Dortmund0,875Leipzig0,880Bonn0,889 Half of the total traffic volume is handled by only 3.2% of the road meters. 50% of all road meters have only 0.2% of the total traffic volume. 80% of total traffic volume is concentrated on no more than 10% of all road meters.

10 10 Institute for Economics and Traffic Chair for Traffic Modeling and Econometrics Stefan Lämmer traffic@stefanlaemmer.de 10 Cellular Structures ► Distribution of cell-degrees (number of neighboring cells) Crack patterns Dragon fly wings Road networks Honey combs

11 11 Institute for Economics and Traffic Chair for Traffic Modeling and Econometrics Stefan Lämmer traffic@stefanlaemmer.de 11 Scaling of Cell Areas Typical values for α Berlin 1.158 Bremen 0.931 Hamburg 0.890 Duisburg 0.924 Munich 1.114 Leipzig 0.926 Cologne 0.922 Nuremberg 0.831 Frankfurt 1.009 Dresden 0.892 Dortmund 0.803 Bochum 0.829 Stuttgart 0.901 Wuppertal 0.883 Essen 0.932 Bielefeld 0.735 Düsseldorf 0.964 Bonn 1.018 Cells with half the cell area are two times more frequent

12 12 Institute for Economics and Traffic Chair for Traffic Modeling and Econometrics Stefan Lämmer traffic@stefanlaemmer.de 12 Distribution of Form Factors ► Form Factor φ: Fraction of the circumscribed circle that is covered by the cell φ=0 … long and narrow φ=1 … compact and round ► High values of Var(φ) imply Irregular network structure, e.g. city has grown over many epochs Difficult to navigate from car driver’s point of view Var(φ) Berlin0,159 Hamburg0,164 Munich0,159 Cologne0,165 Frankfurt0,169 Dortmund0,166 Stuttgart0,170 Essen0,169 Düsseldorf0,175 Bremen0,166 Duisburg0,169 Leipzig0,153 Nuremberg0,172 Dresden0,156 Bochum0,171 Wuppertal0,162 Bielefeld0,161 Bonn0,173

13 13 Institute for Economics and Traffic Chair for Traffic Modeling and Econometrics Stefan Lämmer traffic@stefanlaemmer.de 13 Conclusions ► Scaling of neighborhood sizes Fast roads let neighborhoods grow fast Effective dimension of urban space is significantly higher than two (although mainly planar) ► Scaling of traffic flow Estimation of traffic based on betweenness centrality Power-law distribution of traffic flow on nodes We found quantitative measures to characterize concentration of traffic ► Scaling of cell areas Power-law distribution of cell areas We found quantitative measures to characterize irregularity of cellular structure

14 http://www.helbing.org MOVING THE WORLD. Traffic Science Faculty "Friedrich List" Institute for Economics and Traffic Chair for Traffic Modelling and Econometrics Dresden University of Technology, Germany Stefan Lämmer traffic@stefanlaemmer.de Scaling Laws in the Spatial Structure of Urban Road Networks Physica A 363(1) 89-95

Download ppt "MOVING THE WORLD. Traffic Science Faculty "Friedrich List" Institute for Economics and Traffic Chair for Traffic Modelling and Econometrics."

Similar presentations

Network analysis Sushmita Roy BMI/CS 576

Network analysis Sushmita Roy BMI/CS 576
Maglev‘2006, Dresden, September 13-15 Why Maglev? Dr. rer. pol Michael Raschbichler Transrapid in the context of urban and regional transport planning.

Maglev‘2006, Dresden, September Why Maglev? Dr. rer. pol Michael Raschbichler Transrapid in the context of urban and regional transport planning.
Emergence of Scaling in Random Networks Albert-Laszlo Barabsi & Reka Albert.

Emergence of Scaling in Random Networks Albert-Laszlo Barabsi & Reka Albert.
Analysis and Modeling of Social Networks Foudalis Ilias.

Analysis and Modeling of Social Networks Foudalis Ilias.
VL Netzwerke, WS 2007/08 Edda Klipp 1 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Networks in Metabolism.

VL Netzwerke, WS 2007/08 Edda Klipp 1 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Networks in Metabolism.
Information Networks Small World Networks Lecture 5.

Information Networks Small World Networks Lecture 5.
CONNECTIVITY “The connectivity of a network may be defined as the degree of completeness of the links between nodes” (Robinson and Bamford, 1978).

CONNECTIVITY “The connectivity of a network may be defined as the degree of completeness of the links between nodes” (Robinson and Bamford, 1978).
CS 599: Social Media Analysis University of Southern California1 The Basics of Network Analysis Kristina Lerman University of Southern California.

CS 599: Social Media Analysis University of Southern California1 The Basics of Network Analysis Kristina Lerman University of Southern California.
Weighted networks: analysis, modeling A. Barrat, LPT, Université Paris-Sud, France M. Barthélemy (CEA, France) R. Pastor-Satorras (Barcelona, Spain) A.

Weighted networks: analysis, modeling A. Barrat, LPT, Université Paris-Sud, France M. Barthélemy (CEA, France) R. Pastor-Satorras (Barcelona, Spain) A.
Emergence of Scaling in Random Networks Barabasi & Albert Science, 1999 Routing map of the internet

Emergence of Scaling in Random Networks Barabasi & Albert Science, 1999 Routing map of the internet
Network Models Social Media Mining. 2 Measures and Metrics 2 Social Media Mining Network Models Why should I use network models? In may 2011, Facebook.

Network Models Social Media Mining. 2 Measures and Metrics 2 Social Media Mining Network Models Why should I use network models? In may 2011, Facebook.
Scale-free networks Péter Kómár Statistical physics seminar 07/10/2008.

Scale-free networks Péter Kómár Statistical physics seminar 07/10/2008.
The impact of network density, travel and location patterns on regional road network vulnerability Erik Jenelius Lars-Göran Mattsson Div. of Transport.

The impact of network density, travel and location patterns on regional road network vulnerability Erik Jenelius Lars-Göran Mattsson Div. of Transport.
Lecture 23 (mini-lecture): A Brief Introduction to Network Analysis Parts of the Network Analysis section of this lecture were borrowed from a UC Berkeley.

Lecture 23 (mini-lecture): A Brief Introduction to Network Analysis Parts of the Network Analysis section of this lecture were borrowed from a UC Berkeley.
Identifying Patterns in Road Networks Topographic Data and Maps Henri Lahtinen Arto Majoinen.

Identifying Patterns in Road Networks Topographic Data and Maps Henri Lahtinen Arto Majoinen.
CS 728 Lecture 4 It’s a Small World on the Web. Small World Networks It is a ‘small world’ after all –Billions of people on Earth, yet every pair separated.

CS 728 Lecture 4 It’s a Small World on the Web. Small World Networks It is a ‘small world’ after all –Billions of people on Earth, yet every pair separated.
Peer-to-Peer and Grid Computing Exercise Session 3 (TUD Student Use Only) ‏

Peer-to-Peer and Grid Computing Exercise Session 3 (TUD Student Use Only) ‏
Global topological properties of biological networks.

Global topological properties of biological networks.
Small World Networks Somsubhra Sharangi Computing Science, Simon Fraser University.

Small World Networks Somsubhra Sharangi Computing Science, Simon Fraser University.
Complex Networks Structure and Dynamics Ying-Cheng Lai Department of Mathematics and Statistics Department of Electrical Engineering Arizona State University.

Complex Networks Structure and Dynamics Ying-Cheng Lai Department of Mathematics and Statistics Department of Electrical Engineering Arizona State University.

Similar presentations

Ads by Google