Dynamics of Variable Speed Limit and Driver Information System Surrounding a Bottleneck on a German

1
Dynamics of Variable Speed Limit and Driver
Information System Surrounding a Bottleneck on a
German Autobahn
Steven Boice Robert L. Bertini Portland State
University
Klaus Bogenberger BMW Group
February 9th, 2006 TransNow Student
Conference Oregon State University
2
Outline

• Objectives
• Methodology
• Bottleneck Diagnosis
• Variable Speed Limit and Driver Information
  System Dynamics
• Findings

3
Bottleneck Definition

• An active bottleneck is a restriction that
  separates upstream queued traffic from downstream
  un-queued traffic
• An active bottleneck is de-activated when there
  is either a decrease in demand or when a queue
  spills back from a downstream bottleneck

4
Objectives

• Empirical analysis of reproducible bottleneck
  activations using inductive loop detector data
  from Autobahn 9 (A9)
• Comparison and fusion of unique historical data
  sets
• Variable Speed Limit
• Traveler Information
• Floating Car
• Weather

5
Study Area Autobahn 9 (A9) Kilometers 512-528
6
Data Overview

• Days
• June 24-30, 2002 July 1-8, 2002
• May 21, 2003
• Data
• Inductive Loop Detector
• 17 stations northbound / 16 stations southbound
• 1-minute aggregation
• Vehicle count and time-averaged speed by lane and
  vehicle type
• Limited data for freeway ramps
• Variable Message Sign
• 5 northbound message signs / 10 southbound
  message signs
• Variable speed limit
• Traveler information (warnings/prohibitions)
• Floating Car
• Trajectory in time and space for probe vehicle
• Weather

7
Methodology Analysis Tools

• Transformed cumulative curves (e.g. Newell,
  Cassidy Windover)
• Vehicle count
• Time-averaged speed
• Transformations to heighten visual resolution
• Oblique axis
• Horizontal shift with vehicle conservation
• Retain highest level of resolution (one-minute)
• Identify bottleneck activations and
  deactivations.

8
Methodology Transformed Cumulative Curves
N(x,t)
9
Methodology Transformed Cumulative Curves
N(x,t)
q05180 vph
N(x,t)-q0t
10
Methodology Transformed Cumulative Curves
Time
1400
1415
1430
1445
1500
1515
1530
1545
1600
11,000
10,000
9,000
8,000
N(x,t)-q0t
7,000
6,000
5,000
N(x,t) Cumulative Count
4,000
3,000
2,000
1,000
0
11
Methodology Transformed Cumulative Curves
Time
1400
1415
1430
1445
1500
1515
1530
1545
1600
N(x,t)-q0t
N(x,t) Cumulative Count
12
Methodology Analysis Tools

• Time Space Diagram
• Speed Contour Plot

13
Methodology Analysis Tools

• Queuing Diagram

14
Methodology Analysis Tools







N(x1,t)

Travel Direction

N(xj,t)


x

x

Excess Accumulation
1
2
k

N(x2,t)
Excess Travel TimeDelay


Time, t
t2
15
Bottleneck Diagnosis Speed Contour June 27,
2002 A9 Northbound (1500-2000)
Un-queued
Queued
16
Bottleneck DiagnosisFloating Car Trajectory
17
Bottleneck Diagnosis Queuing Diagram June 27,
2002 Northbound
1524_at_ Station 420
1521_at_ Station 390
1521_at_ Station 380
Flow Reduction _at_380
N(x,t)-q0t', q05170 veh/hr
18
Bottleneck Diagnosis Triggers
19
Bottleneck Summary

• Northbound
• 11 activations over 6 days between stations 380
  and 390
• 5 isolated bottlenecks with flow drop activated
  between stations 380 and 390
• 6 non-isolated bottlenecks activated between
  stations 380 and 390
• Southbound
• 3 activations over 4 days between stations 57 and
  53
• 3 isolated bottlenecks with flow drop activated
  between stations 57 and 53

20
Bottleneck Summary Isolated
21
Bottleneck Characteristics Isolated

• Pre-Queue Flow
• Consistent from day to day
• 5728 vph northbound (3 lanes)
• 4512 vph southbound (2 lanes)
• Comparable to A5 (5193 vph w/3 lanes)
• Discharge Flow
• Consistent from day to day
• 5401 vph northbound (3 lanes)
• 4075 vph southbound (2 lanes)
• Comparable to A5 (4985 vph w/3 lanes)
• Shockwave Speed
• Consistent from day to day
• Range (-)6-20 km/hr
• Average (-)14 km/hr
• Comparable to A5 average (-)18 km/hr

22
VSL and Driver Information Dynamics Speed
Contour June 24, 2002 A9 Southbound (500-1200)
Shock Speed (-) 14 km/hr
23
VSL and Driver Information Dynamics Variable
Speed Limit
24
VSL and Driver Information Dynamics
BN
25
VSL and Driver Information Dynamics
Warnings/Prohibitions
26
VSL and Driver Information Dynamics Textual
Information
27
Summary of Findings

• Transformed cumulative vehicle count and speed
  curves can reveal resolution necessary to
  identify bottleneck activation
• Backward moving shocks have a consistent speed
• (-)6-20 km/hr
• Bottleneck discharge flow found to be lower than
  pre-queue flow prior to bottleneck activation
• Supports two capacity theory
• (-)8
• Evidence supports main triggers of bottleneck
  activation were
• Increase in flow at both on and off ramps near
  bottleneck location
• Increase in truck traffic in median lane
• Increase in systematic truck lane changing
• Trucks were found to disregard the no passing
  prohibition and found to travel on average 10-20
  km/hr greater than mandatory 80 km/hr

28
Summary of Findings

• Evidence of correlation between actual traffic
  conditions and
• Variable speed limit
• Traveler information

29
Contribution

• Improved travel time estimation and forecasting
• Traffic management
• Traveler information
• Driver assistance systems
• Contribute to improved traffic flow models and
  freeway operational strategies
• Opportunity to compare with previous findings
  using data from German freeways (bottleneck
  capacity)

30
Continued Analysis

• Before/After comparison of variable speed limit
  system to determine system benefits
• Continued empirical analysis and comparison with
  measurements on freeways in the U.S., Canada, and
  United Kingdom

31
Acknowledgements

• Dr. Robert Bertini, Dr. Soyoung Ahn, and Steve
  Hansen
• Dr. Klaus Bogenberger, Dr. Georg Lerner, Dr.
  Hartmut Keller, and the BMW Group
• Christian Mayr and Frank Frischeisen of
  Autobahndirektion Südbayern

32
Thank you for your attention.
http//www.its.pdx.edu/