Welcome to Todays T3 Webinar Signal Timing Optimization

Welcome to Todays T3 Webinar Signal Timing Optimization

Welcome to Todays T3 Webinar Signal Timing Optimization Using Connected Vehicle Technology August 13, 2018 1:00 pm Listen through your speakers/VoIP If audio cannot be heard 5 minutes prior to the start of this webinar, please dial into the webinar using the teleconference number provided onscreen. T3 Webinars are sponsored by ITS Professional Capacity Building Program, ITS Joint Program Office, U.S. Department of Transportation Signal Timing Optimization Using Connected Vehicle Technology Xiao (Joyce) Liang PhD Candidate, Penn State University [email protected] USDOT T3e Webinar August 13, 2018 Traffic Operations Group at Penn State Vikash Gayah

Ilgin Guler Murat Bayrak Network wide infrastructure improvements Josh Killian Predicting transit dwell times Joyce Liang Signal control utilizing CAVs Kan Wu Transit signal priority at intersections and arterials Guanhao Xu Macroscopic traffic flow models Rebecca Yocum

Macroscopic traffic flow models Yinghai Yu Weather impacts on traffic flow Zhengyao Yu Resilience of urban street network configurations Urban Mobility, Optimization and Control, Statistical Modeling of Transportation Systems Vikash Gayah Large-scale transportation system behavior Prediction of: safety performance transit system evolution mode choice

driver behavior Statistical modeling of transportation data How can we model behavior of large-scale transportation systems? Design, control and management of safe, efficient and reliable transportation systems How can we optimize traffic control at signalized intersections accounting for various vehicle types? Optimization frameworks for transportation

systems Organization of street networks How does street layout impact efficiency, environmental impacts, safety, and network reliability? Transportation Systems and Infrastructure Management Public transportation systems on urban networks Ilgin Guler Public Transportation Making better management decisions Can we optimize the maintenance and repair decisions for infrastructure?

What are the effects of different types of public transportation systems on general traffic on a network? Infrastructure Management Managing Transportation Infrastructure Car to car communications Can we use new technology to improve traffic operations? Traffic Management and Control Multi-modal Traffic Interactions Bus priority How can traffic operations at a signalized

intersection be optimized considering both cars and buses? Traffic Safety Multi-modal safety Can we define different metrics for measuring safety of multiple modes? Utilizing Connected Vehicle Technology to Improve Signal Operations Objective: To optimize traffic operations (minimize delay and number of stopping maneuvers) at an intersection by using information from connected vehicles such as the position and speed of individual vehicles and autonomous vehicle presence. Strategy: Joint optimization of signal timing using CV information and trajectory/speed guidance to AVs http://www.dot.state.fl.us/trafficoperations/ITS/Projects_Deploy/CV/ Connected_Vehicles.shtm

Outline Background introduction Proposed method Results and discussion Extension to the proposed method 08/13/18 USDOT T3e webinar 7 Introduction Three traditional traffic signal control methods: Fixed-time Actuated Adaptive 08/13/18 USDOT T3e webinar 8

Introduction Connected vehicles: Share information with infrastructure. Autonomous vehicles: http://www.dot.state.fl.us/trafficoperations/ITS/Projects_Deploy/CV/ Connected_Vehicles.shtm Connected vehicles Self-driving Trajectories can be controlled. https://www.google.com/url?sa=i&rct=j&q=&esrc=s&source=images&cd=&cad=rja&uact=8&ved=2ahUKEwjK2smt1tzcAhVyGDQIHdtLCJgQjRx6BAgBEAU&url=http%3A%2F%2Ffitnflash.com %2Fhow-driverless-cars-see-the-world-around-them%2Fself-driving-car-illustration-1521499338405-facebookjumbo-png %2F&psig=AOvVaw0MhUw5UeG6fwY5AxA3hWYo&ust=1533790800080574 08/13/18 USDOT T3e webinar 9 Introduction

Connected vehicle technology can be used to: Modify trajectory of fully autonomous vehicles (safety application) Li and Wang, 2006; Zohdy and Rakha, 2012; Lee and Park, 2012 Optimize phases (cycle length and green splits) of a signal (operations application) Grandinescu et al., 2007; He et al., 2012; Lee et al., 2013; Goodall et al. 2013 Optimize vehicle discharge sequence (operations application) Dresner and Stone, 2004; Wu et al. 2007; Cai et al., 2012 08/13/18 USDOT T3e webinar 10

Goal To optimize traffic operations at an intersection Minimize average vehicle delay Minimize average number of vehicle stops Method: Utilizing connected vehicle information, such as the position and speed of individual vehicles, to optimize signal plan Trajectory design of autonomous vehicles. 08/13/18 USDOT T3e webinar 11 Signal Control Algorithm Three types of vehicles are considered: Basic intersection configuration Traditional vehicles, Connected but non-autonomous vehicles

(connected vehicles), and Autonomous vehicles Inputs: Information obtained from connected vehicles: Zone of interest 1. 2. 3. 08/13/18 The time it enters the zone of interest The distance from the intersection at which it comes to a stop (if a queue exists) Real-time location, speed, acceleration USDOT T3e webinar 12

Signal Control Algorithm 08/13/18 USDOT T3e webinar 13 Step 1: Traditional Vehicle and Platoon Identification Traditional vehicles are identified if a CAV stops behind it. Cars are platooned based on Minimum spacing (stopped vehicles) Minimum headway (moving vehicles) 08/13/18 USDOT T3e webinar 14 Step 1: Traditional Vehicle and Platoon Identification Time t1 Time t3

Time t2 Light blue: connected but not autonomous Yellow: traditional vehicles. Rectangle: No platoon Oval: Platoon 08/13/18 USDOT T3e webinar 15 Step 2: Solution Method to Identify Optimum Signal Phasing Plan Estimate delays of departure sequences to identify the optimal departure sequence that will result in minimum delay Enumeration method: simply identifies all possible combinations of platoon departure sequences 08/13/18 USDOT T3e webinar

16 Step 2: Solution Method to Identify Optimum Signal Phasing Plan 3 1 2 4 Signal phasing and timing plan: 2 08/13/18 Enumeration: 6 possible departure combinations considered: 1,2,3,4 1,2,4,3 1,3,2,4 2,1,3,4

2,1,4,3 2,4,1,3 1+3 USDOT T3e webinar 4 17 Step 3: Longitudinal Trajectory Guidance Input: Optimum signal phase and signal timing plan Expected departure time Modify car trajectories to: Let vehicles pass the intersection at a specific time. Slow down to avoid stopping during red time, or Speed up to not waste green time 08/13/18 USDOT T3e webinar 18 Step 3: Longitudinal Trajectory Guidance

Accounts for realistic acceleration or deceleration of cars 08/13/18 USDOT T3e webinar 19 Step 3: Longitudinal Trajectory Guidance Time t1 Time t2 Time t3 Dark blue: autonomous vehicles. Light blue: connected but not autonomous Yellow: conventional vehicles. Rectangle: No platoon Oval: Platoon 08/13/18

USDOT T3e webinar 20 Result: Benefits of Platooning Computation time [ms] 50 3500 3000 40 2500 2000 30 1500 20

1000 500 10 1.5 2 2.5 3 3.5 4 4.5 5 Critical headway [s] 08/13/18

USDOT T3e webinar 21 Result: Benefits of Platooning Computation time [ms] 50 3500 3000 40 2500 2000 30 1500 20 1000

Average delay [s] 50 500 10 1.5 A critical headway of 2.5 seconds and critical Computation time significantly reduces as spacing of 15 meters is chosen since it provides more cars are platooned together significant computational efficiency without much change to average delay or stop 10.6 2 2.5 3

3.5 4 4.5 40 5 10.5 Critical headway [s] Average delay increases slightly as more cars are platooned together 50 30 10.4 20

10.3 10.2 10 Average number of stops 1.5 0.05 2 2.5 3 3.5 4 4.5 5

Critical headway [s] 40 0.04 30 0.03 20 0.02 10 Average number of stops increase as more cars are platooned together but the magnitude of increase is small. 0.01 1.5 2

2.5 3 3.5 4 4.5 5 Critical headway [s] 08/13/18 USDOT T3e webinar 22 Result: Sensitivity of algorithm to penetration ratio Computation time significantly increases as more cars are connected, and also as more are autonomous

Average delay reduces with more information, but marginal benefits after 40% connected ratio are very small Average number of stops decreases as more cars are autonomous 08/13/18 USDOT T3e webinar 23 Extensions: 1. More complicated intersection configurations 08/13/18 USDOT T3e webinar 24 Extensions: 1. More complicated intersection configurations

08/13/18 USDOT T3e webinar 25 Extensions: 2. More phase options 08/13/18 USDOT T3e webinar 26 Extensions: 3. Other vehicle types Human-driven vehicles that receive guidance Reaction time Speed acceptance Execution error 08/13/18 USDOT T3e webinar

27 Conclusions Methodology can: Identify non-connected vehicles using data from CVs Group vehicle into platoons that naturally discharge together Optimize platoon discharge sequence to minimize vehicle delays Alter vehicle trajectories to minimize number of stops Methodology used to examine trade-off between computational complexity and operational performance 08/13/18 USDOT T3e webinar 28 Conclusions

Platoon-based algorithm provides significant reductions in computational time over vehicle-based methods with minimal changes to average vehicle delay or number of stopping maneuvers. The algorithm provides larger and more significant operational benefits as the penetration rate of CAVs increases. However, the marginal benefits are much smaller after the fleet is composed of 40% CAVs, since platoons are well-identified at this penetration rate. As the penetration rate of AVs increases, trajectory guidance for them can provide significant reductions in number of stops. The algorithm can be extended to more complicated signal configuration and more phase options. Future work: Adding pedestrians, buses, emergency vehicles, etc. 08/13/18 USDOT T3e webinar 29 Thanks! Xiao (Joyce) Liang [email protected] 08/13/18

USDOT T3e webinar 30

Recently Viewed Presentations

  • Managing Behavioural Disturbance Dr Dan Mosler, Consultant Psychiatrist

    Managing Behavioural Disturbance Dr Dan Mosler, Consultant Psychiatrist

    Managing Behavioural Disturbance Dr Dan Mosler, Consultant Psychiatrist Dr Marty Downs, Psychiatric Registrar Verbal de-escalation techniques Use the "three F's" (feel, felt, found) I understand how you could feel that way.
  • Match Officials Round of 16, First Leg CD

    Match Officials Round of 16, First Leg CD

    Among the new arrivals to the 2017 Torneo Apertura champions are FW Leandro Rodriguez, FW Yuaycell Wright, MF Rafael Rodriguez, MF Javier Liendo, GK Carlos Mendez, DF Asdrubal Gibbons and DF EdderMonguio.
  • TeamSCO Services Webinar May 7, 2003 Kerri Wallach,

    TeamSCO Services Webinar May 7, 2003 Kerri Wallach,

    Services COGS (includes Boston) Web Services. Web Services (includes CTO) FY 03. 10) FY 2002 and FY 2003. Target vs. Roll Up (1st Draft) FY 2002-2003 Proposed Target. Fcst. FY 2002-2003 Roll Up. FY 2002 - 2003 Operating Plan Budget-OS...
  • Singing Social Friction Community Building Through Antihegemonic Performances

    Singing Social Friction Community Building Through Antihegemonic Performances

    i.e. „monodic" (without accompanying harmony, either instrumental or vocal) Claim that singing with instrumental accompaniment is a modern innovation...is debated (singers used whatever was at hand; the collectors preferred to collect the songs unaccompanied)
  • Religious Factors in Modern Economies

    Religious Factors in Modern Economies

    - Elder Dallin H. Oaks, from address given at BYU-I, "Religious Freedom", emphasis added CHALLENGE: To properly define its meaning PROBLEM: The meaning of religious freedom Why is there a battle over the meaning of religious freedom? Few contest religious...
  • Hra opakování učiva ZŠ 1

    Hra opakování učiva ZŠ 1

    Vypočítejte obsah původní podložky. Zpět Správná odpověď: Pro správnou odpověď klikni na smajlíka. 10205 cm2 Otázka číslo 17 Je dána kružnice určená středem S. Na kružnici leží body A; C; B v tomto pořadí. Úhel ASC měří 780, úhel CSB...
  • Gitchi Gummi Soccer Club Brief History  Gitchi Gummi

    Gitchi Gummi Soccer Club Brief History Gitchi Gummi

    Brief History Gitchi Gummi Soccer Club was founded in 1991 By Greg Cane, Nic Bacigalupo, Shelley Hedtke, Alex Giuliani, and others. They wanted to form teams from around the Duluth area that were capable of playing at the top level...
  • Scientific Method, Theories & Laws

    Scientific Method, Theories & Laws

    Scientific Law. A scientific law is a statement based on repeated experimental observations that describes some aspects of the universe. A scientific law always applies under the same conditions, and implies that there is a causal relationship involving its elements.