Traffic Lights Animated Templateacademic.uprm.edu/amfigueroa/Presentations/2011... · 3/23/2012 22...
Transcript of Traffic Lights Animated Templateacademic.uprm.edu/amfigueroa/Presentations/2011... · 3/23/2012 22...
3/23/2012
1
Agenda
National issues in traffic signal
control
Every Day Counts and Adaptive
Signal Control Technologies
Existing Algorithms
ASCT Implementation in Puerto Rico
Corridor Optimization with ACS-Lite
Page 2
3/23/2012
2
2007 National Traffic Signal Report Card
Improper traffic signal timing
accounts for 5-10% of all traffic delay
on major roadways
Incorrectly functioning traffic sensors
do not serve all vehicles and
pedestrians equitably.
Drivers must wait through more than
one green signal at an intersection,
causing long queues and clogged
intersections.
Report Card Findings
Topic Area Existing conditions / problems
Management
Score: D-
• Philosophy for signal operation not documented or shared
• Annual reviews of major roadways are rarely conducted
• No established business plan for transportation operations
with clearly defined performance measures and goals
Signal Operation in
Coordinated
Systems
Score: D
• Traffic signal timing is rarely reviewed
• Signal technicians are not current on the use of modern
software
• Timing plans are not in place for emergencies and special
events
Signal Timing
Practices Score: C
• Intersection operations are infrequently checked in the field
to accommodate changing traffic conditions.
Traffic Monitoring &
Data Collection
Score: F
• Real-time traffic data are seldom available to the traveling
public for information and route planning
• Few, if any, quality checks for traffic monitoring and
collection systems
Page 4
3/23/2012
3
0
500
1000
1500
2000
2500
3000
3500
4000
12:00
AM
2:00
AM
4:00
AM
6:00
AM
8:00
AM
10:00
AM
12:00
PM
2:00
PM
4:00
PM
6:00
PM
8:00
PM
10:00
PM
Time of Day
VP
H (
veh
icle
s p
er
ho
ur)
AM
PEAK 15 Min
Opportunities to Improve Traditional
Traffic Signal Timing
Traditional signal timing process emerged
between 1940’ss & 1960’s
Limited signal timing plans
Variable & unpredictable traffic conditions
Every Day Counts “Every Day Counts is designed to identify and deploy
innovation aimed at shortening project delivery, enhancing
the safety of roadways, and protecting the environment”
Víctor Méndez, FHWA Administrator
Every Day Counts
Make FHWA a greener Agency and reduce their
carbon footprint
Accelerating Technology and Innovation
Deployment
Warm Mix Asphalt
Prefabricated Bridge Elements and Systems
Adaptive Signal Control Technology
Safety Edge
Geosynthetic Reinforced Soil
Shortening Project Delivery
Shortening Project Delivery Toolkit
Accelerated Project Delivery
Methods Adaptive Signal
Control Technology
3/23/2012
4
Adaptive Signal Control Technology
What is ASCT?
Sensors monitor traffic
Software evaluates
performance
Timing updated if necessary
Process repeats
What is ASCT?
Sensors monitor traffic
Software compares to baseline timing plan
Timing change if necessary
Process repeats
Sensors
•Type
•Placement
•Effective
•Well-maintained
3/23/2012
5
What is ASCT?
Sensors monitor traffic
Software compares to baseline timing plan
Timing change if necessary
Process repeats
Evaluate Performance
•Data
•Timing Parameters
•Algorithms & Models
•Frequency of Analysis
What is ASCT?
Sensors monitor traffic
Software compares to baseline timing plan
Timing Updates if necessary
Process repeats
Update Timing
•Uncongested
•Splits
•Offset
•Cycle PM Peak Period Demand
250
300
350
400
450
500
3:0
0 - 3:1
5
3:1
5 - 3:3
0
3:3
0 - 3:4
5
3:4
5 - 4:0
0
4:0
0 - 4:1
5
4:1
5 - 4:3
0
4:3
0 - 4:4
5
4:4
5 - 5:0
0
5:0
0 - 5:1
5
5:1
5 - 5:3
0
5:3
0 - 5:4
5
5:4
5 - 6:0
0
6:0
0 - 6:1
5
6:1
5 - 6:3
0
6:3
0 - 6:4
5
6:4
5 - 7:0
0
Time
De
ma
nd
(V
PH
) SB
NBLT
WBLT
WBEB
NB
3/23/2012
6
What is ASCT?
Sensors monitor traffic
Software compares to baseline timing plan
Timing change if necessary
Process repeats
Continuous
Monitoring
•24/7
•Weekdays, Weekends
•Holidays
•Special Events
•Incidents
Implementation Goals for
Adaptive Signal Control Technologies
Goal 1
By December 2011, ASCT will
be comprehensively evaluated
and demonstrated to
underscore its opportunities
and benefits
Goal 2
By December 2012, ASCT /
EDC tools will be used by 40
agencies
3/23/2012
7
EXISTING ALGORITHMS
Page 13
Existing Algorithms
SCATS
ACS Lite
RHODES
SCOOT
OPAC
InSync Prediction Generators
Prediction Generators
Prediction Generators
Prediction Receiver
3/23/2012
8
System Requirements
Algorithm Controller type Detection Required
Field /Central Based
Algorithm
Field Processor Required
ACS Lite Various legacy controllers/cabiets
Existing stop bar detection by phase. Prefer one or more advance detectors per coordinated phase
Field or central
Yes
SCATS SCATS firmware on EPAC M50 and 2070
Stop bar only by lane Central No
SCOOT SEPAC firmware on EPAC M50 and 2070
stop bar and upstream detection
Central No
RHODES Siemens NextPhase 1.7.6c-2070
stop bar and upstream detection
Field Yes
OPAC Econolite ASC2, ASC3; Other NTCIP controllers on way
stop bar and upstream detection
Field or central
Yes
InSync Various - can work with most controller firmware
Video stop bar only Field Yes
Cost Comparison
Algorithm Detection cost
Aditional hardware
cost
Average software
developer cost
Average total cost per
intersection
ACS Lite $35,000 $0 $4,500 $39,500
SCATS $20,000 $0 $17,000 $37,000
SCOOT $50,000 $0 $20,000 $70,000
RHODES $50,000 $50,000 $50,000 $150,000
OPAC $50,000 $1,200 $3,000 $54,200
InSync N/A N/A N/A $78,000
3/23/2012
9
ASCT Deployment Status
SCATS
SCOOT
ACSLite
InSync
RHODES
OPAC
LA ATCS
State with Active ASCT
State with Pending ASCT
Traffic Signal Modernization Project and
Adaptive Signal Control in Puerto Rico
Update traffic signal equipment with Econolite ASC/3
NEMA TS2/NTCIP Actuated Controllers
Replacement of wire detection loops with Autoscope
Rack Vision Terra video detection system
Installation of radar sensors for traffic data collection
(volume, vehicle classification, speeds) in segments
Installation of video cameras for traffic and incident
monitoring
Installation of wireless communication infrastructure in
PR-2 corridor
Establish a local Traffic Management Center at
Mayaguez HTA Regional Office 19
3/23/2012
10
Econolite Signal Control
Equipment
20
Autoscope Rack Vision Terra
Video Detection System
21
3/23/2012
11
Traffic Monitoring, Data Collection
and Communication Equipment
22
MMA TMC and
Centracs Signal Management System
Provides a platform
for monitoring and
control of all signals
in the corridor.
ASCT can be
incorporated
to optimize signal
cycles and offsets.
Provides reports of flow
by lane in different
periods of time, vehicle
type, lane utilization and
space-time diagrams.
3/23/2012
12
Page 24
Typical ACS Lite System
CORRIDOR OPTIMIZATION WITH
ACS-LITE
1. Split tuning algorithm
2. Offset tuning algorithm
Page 25
3/23/2012
13
Collect Data
1. SPLIT TUNING ALGORITHM
Ø3
Ø7
Ø4
Ø8
Ø4
Ø7
Ø3
Ø3 Ø8
Ø5
Ø2
Ø7 Ø4
Ø1
Ø6
Ø5
Ø2
Ø5
Ø2
Correlate data to signal
phasing
3/23/2012
14
Averaging
Ø3 Cycle 1
Ø3 Cycle 2
Ø3 Cycle n-1
Ø3 Cycle n
Ø3 Phase
Utilization
Perform Analysis
Ø3 Ø8
Ø5
Ø2
Ø7 Ø4
Ø1
Ø6
Implement phase split
adjustments
3/23/2012
15
Volume and occupancy data is collected by
the advanced detectors usually 250 feet or
more upstream of the stop bar.
Collect data from
advance detectors
on coordinated
approaches
2. OFFSET TUNING ALGORITHM
Phase State
Volume
Occupancy
Green Interval
Vehicles arriving during
yellow and red intervals
Statistical flow profile using the volume
and occupancy data.
Develop a Statistical Flow Profile
correlated to the phase state
3/23/2012
16
Phase State
Volume
Occupancy
Green Interval
Red Arrow points out vehicles arriving
before the start of green. Shifting the
offset to the left will match the green
interval to a greater percentage of arrivals.
Perform analysis to capture the most
arriving flow during the green interval
Phase State
Volume
Occupancy
Green Interval
Implement offset adjustment
3/23/2012
17
ACS-Lite Idaho Case Study
• US-95 section in Coeur d’Alene,
Idaho with four signalized
intersections
• Evaluation conducted using PM
peak traffic conditions
• Traffic volumes, turning
movements and network
geometry information were taken
from an existing Synchro model
• Signal timing plans were
extracted from field controllers
Corridor Travel Time Comparison
3/23/2012
18
ACS-Lite Fulton Co. Case Study
Cascade Road section in
Atlanta, GA with 5 signalized
intersections
Deployment of ACS-Lite
included:
Upgrade controller firmware
Convert controller database
Set up detectors
System Integration Page 36
Side Street Queue Length Comparison
Page 37
3/23/2012
19
ACS-Lite Morgantown, VA Case Study
Highway corridor with 19
signalized intersections
Vehicle counts were obtained
from 14 hour field
observations
Three TOD timing plans were
developed in Synchro
Page 38
Average Delay Comparison
Page 39
3/23/2012
20
Corridor Travel Times
Page 40
Arrivals on Green w/o ACS-Lite
Page 41
3/23/2012
21
Arrivals on Green w/ ACS-Lite
Page 42
Conclusions
Opportunities for improvements in traffic signal
operations with ACST are
Non-saturated corridors
Irregular traffic patterns / diverge from simulated
conditions
Improvements in travel time and arrivals on green,
individual intersection delay might increase
Corridor selection guidelines for ASCT
implementation are not developed
Issues with high left-turn movements and
pedestrian flows & preemption operation
Page 43
3/23/2012
22
References
FHWA. 2010. Every Day Counts Summit
Presentations. Atlanta, GA.
Transpo Group. 2007. US-95 ACS-Lite System
Evaluation. Idaho.
Wang, J. et al. Evaluation of ACS Lite Adaptive
Control using Sensys Arterial Travel Time Data
Day, C. et al. 2011. Adaptive Signal Control
Performance Measures: A Central System-in-the-
Loop Simulation Case Study.
Page 44