Transcript Signal - University of Nevada, Reno
A Probabilistic Approach Determining When to Turn on/off Signal Coordination
Rasool Andalibian Center for Advanced Transportation Education and Research April 2014
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Outline
Background and Problem Statement Signal Coordination: Common Practice Stop Probabilistic Model Model Outputs Summary and Conclusions
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Problem Statement
Major signalized arterials are generally coordinated during peak periods.
They run free (actuated) during non-peak periods. Traffic demand level is a key element to consider. At what demand level signal coordination is warranted?
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Signal Coordination Strategy
Signal Timing Manual : intersections in close proximity with large amount of traffic on coordinated street.
MUTCD : Traffic signal within 0.5 mile of each other FHWA : I ntersections close together (i.e., within ¾ mile): advantageous to coordinate them. At greater distances (over ¾ mile), consider the traffic volumes and potential for platoons
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Research Objectives
Develop a probabilistic model that predicts the number of stops for non-coordinated signalized arterials.
Develop # stop thresholds using the model that can guide engineers to decide when signals should be coordinated.
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Previous Work
TRB 2013: Performance Assessment on Non-coordinated Signalized Arterials and Guidelines for Signal Coordination
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Stop Prediction Model
Signal are running free.
Min-recall placed on major arterial.
Probability of stop is independent.
• Probability of stop: 𝑃 𝑖 = 𝐶 𝑖 Probability of hitting green is: 1 − 𝐶 𝑖 Traffic is under-saturated.
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Probabilistic Model: Basic Equations
i
g i ( a ) r i ( a ) P g i
(
a
)
g i
(
a
) /
C
(
a
)
P r i
(
a
)
r i
(
a
) /
C
(
a
)
C
(
a
)
g i
(
a
)
r i
(
a
)
P g i
(
a
)
P r i
(
a
) 1
i
= direction of travel
a
= intersection index
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Probability of Making Stops
Probability of making
x
stops out of
n
intersections: 𝑛 𝑛 𝑃𝑟 𝑥 = ( 𝑝 𝑖 𝑟 {1 − 𝑝 𝑗 𝑟 }) 𝑍 𝑖=1 𝑖≠𝑗 𝑗=1 𝑗≠𝑖 An approximation to the above equation is: 𝑃𝑟 𝑥 = 𝑛 𝑥 ( 𝑟 𝑝 𝑟 )
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Stop Probability: Stop Example
#1 #2 #3
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Stop Probability: Stop Example
Probability of making 1 stop
Pr 1 = 𝑝 1 𝑟 . 1 − 𝑝 𝑟 2 . (1 − 𝑝 3 𝑟 ) Pr 1 = (1 − 𝑝 1 𝑟 ). 𝑝 2 𝑟 . (1 − 𝑝 3 𝑟 ) Pr 1 = 1 − 𝑝 1 𝑟 . (1 − 𝑝 𝑟 2 ) . 𝑝 𝑟 3
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
𝑝 1 𝑟
Stop Probability: Stop Example
Probability of making 1 stop
𝑝 2 𝑟 𝑝 3 𝑟 𝑝 𝑟 𝑃𝑟 1 = 3 1 ( 𝑟 𝑝 𝑟 )
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Probability Distribution of Stops
n=4
0,6 0,4 0,2 0,0
g/C
1,0 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0,8
n=10
0,4 0,2 0,0
g/C
1,0 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0,8 0,6
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Traffic Volume vs. g/C Ratio Intersection Inventory
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Traffic Volume vs. g/C Ratio
Total Volume Distribution
Major
0.50
0.55
0.60
0.65
0.70
Minor
0.50
0.45
0.40
0.35
0.30
Volume Distribution Directionality
Major
0.60 / 0.40
0.70 / 0.30
0.80 / 0.20
Minor
0.55 / 0.45
0.60 / 0.40
0.65 / 0.35
Total entry traffic volume varies from 100 to 5000 vph
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Traffic Volume vs. g/C Ratio
SC_01_A
0.75
0.70
0.65
0.60
0.55
0.50
0.45
0.40
0.35
0 500 1000 1500 2000
Total Traffic Volume
2500 3000
SC_01_B
0.75
0.70
0.65
0.60
0.55
0.50
0.45
0.40
0 500 1000 1500 2000
Total Traffic Volume
2500 3000
SC_02_A
0.75
0.70
0.65
0.60
0.55
0.50
0.45
0.40
0 500 1000 1500 2000
Total Traffic Volume
2500 3000
SC_02_B
0.75
0.70
0.65
0.60
0.55
0.50
0.45
0.40
0 500 1000 1500 2000
Total Traffic Volume
2500 3000
SC_03_A
0.75
0.70
0.65
0.60
0.55
0.50
0.45
0 500 1000 1500 2000
Total Traffic Volume
2500 3000
SC_03_B
0.75
0.70
0.65
0.60
0.55
0.50
0 500 1000 1500 2000
Total Traffic Volume
2500 3000
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Stop Thresholds
Pr 𝑥 > 𝑋 = 𝐾
s.t.
𝑋 = 0.5𝑛 𝑎𝑛𝑑 0.6𝑛 𝐾 = 0.5, 0.6, 0.7, 0.8
𝑛 = 4, 6, 8, 10 𝐼𝑓 𝑛 = 10, 𝑋 = 0.6𝑛, 𝐾 = 0.5
Pr(𝑥 > 6) = 0.5
It is interpreted as 50 percent of drivers will make more than 6 stops.
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Results of Stop Thresholds
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Model Outputs: Recommendation for Signal Coordination
Establishing various stop thresholds results in different level of traffic volumes.
Considering more than
0.5n
and
0.6n
stops with the probability of
0.5
and
0.6
the recommended traffic volume for signal coordination would be:
250 to 350 vphpl
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
ITE Survey
A survey conducted on the ITE Community Website: When Signals are Coordinated • Florida: • San Diego: • Portland: • Sacramento: 250 vphpl 300 vphpl 300 vphpl 350 vphpl
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Summary and Findings
Lack of consistency in traffic demand in signal coordination practice.
This study looks at signal coordination from number of stops standpoint.
A probabilistic stop-base model predicting the distribution of stops.
is developed
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Summary and Findings Cont.
The number of stops is a function of number of intersections and average g/C ratio of all intersections.
An attempt is made to relate actuated g/C ratios and traffic volumes.
Establishing various stop-base thresholds leads to different traffic level for signal coordination.
The author’s threshold is : 50 to 60 percent of drivers making more than 05n and 0.6n stops .
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
Summary and Findings Cont.
The recommended traffic level coordination is 250 to 350 vphpl.
to trigger signal ITE survey shows that the results of this study is compatible with state-of-the-practice.
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014
THANK YOU QUESTION?
“Signals are coordinated according to speed limit thus, NEVER SPEED UP!” Rasool Andalibian
24
C.A.T.E.R
Center for Advanced Transportation Education and Research Class Seminar Spring 2014