ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS AND













































































- Slides: 77
ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS) by Prof. Markos Papageorgiou Dynamic Systems and Simulation Laboratory, Technical University of Crete, Chania, Greece
OUTLINE 1. 2. 3. 4. 5. 6. 7. Introduction Motorway (Freeway) Traffic Control Ramp Metering Route Information and Guidance Variable Speed Limits Urban Signal Control Integrated Urban-Freeway Traffic Control 8. Motorway Traffic Surveillance 9. Other Control Tools ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 2
1. INTRODUCTION Who are we? Dynamic Systems and Simulation Laboratory (DSSL) (~ 15 Professors, Researchers, Ph. D/MSc Students) at the Technical University of Crete also acting occasionally as professional engineers. What is our Background? Automatic Control • Electrical Engineering • Industrial Engineering • Civil Engineering • ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 3
Experience • 30 years of high-quality research in traffic flow modelling and control • Numerous successful implementations • High number of research or implementation contracts: national, EC, companies, authorities worldwide. • October 2008: IEEE Outstanding ITS Research Award ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 4
What We Do Concepts Algorithms Software for traffic flow modelling, surveillance, route guidance, traffic control (freeway + urban) What we Don’t Do Implementation hardware, System integration Our Usual Partners • Road authorities • Local Consultants/University groups • Local System Integrators ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 5
Our Principles • Theoretical soundness: There is nothing more practical than a good theory • Practical Usefulness. • General Applicability, Interoperability • Highest Efficiency: deep understanding • Practicability: As simple as possible, as complex as necessary. ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 6
Automatic Control Loop ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 7
Example: Urban traffic control system ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 8
Minimization of Total Time Spent Maximization of (Early) Exit Rates ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 9
Simple Queuing Systems • Demand > Capacity Queuing • Capacity ≠ f (Queuing) ð Delay depends on D−C only! Water Systems More inflow Higher Pressure Higher Outflow ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 10
Traffic Networks • Congestion degrades the infrastructure (capacity) Local link demand exceeds local capacity Local congestion degrades local capacity Accelerated increase of congestion Further capacity degradation . . . until generalized network congestion although Demand << Nominal network capacity ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 11
Conclusion: Generalized traffic congestion is not only due to high demand. Congested Motorway Networks: Expensive infrastructure capacity not fully available at the only time it is actually needed, i. e. the peak periods! Goal: Operate motorway and urban networks optimally (as controllable systems) ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 12
2. MOTORWAY (FREEWAY) TRAFFIC CONTROL Motorways were originally conceived to provide virtually unlimited mobility to road users, but … ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 13
Man has reached to the moon but … … even ants were taught by evolution to address their transportation problems more efficiently, see I. D. Couzin and N. R. Franks: “Self-organized lane formation and optimized traffic flow in army ants”, Proc. R. Soc. Lond. B (2003) 270, 139– 146 ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 14
Available Motorway Control Measures n Ramp metering (RM): valuable; limited storage space n Variable speed limits (VSL): improved safety; no system improved efficiency n Route guidance (RG): best under incidentcaused congestion n Vehicle-infrastructure integration (VII): promising; emerging ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 15
n Prerequisite for efficient traffic control: Understanding the reasons for infrastructure degradation! n (Latent) Motorway bottleneck location: Capacity upstream > Capacity downstream n Bottlenecks are candidates for congestion appearance: – – – on-ramp merge geometry (lane drop, grade, curvature, tunnel, …) weaving speed limits over-spilling off-ramps ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 16
Two Main Reasons for Motorway Infrastructure Degradation 1. Capacity Drop (CD) CD not well-understood but is deemed to occur due to vehicle acceleration ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 17
2. Blocking of Off-Ramps (BOR) off-ramp flow reduced: vehicles bound for the off-ramp contribute to accelerated congestion increase! ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 18
3. RAMP METERING ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 19
Why Ramp Metering 1 st Answer e. g. qcon = 0. 95 qcap ; qin+d = 1. 2 qcap (veh/h) ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 20
2 nd Answer Note: On-ramp queue should not interfere with surface street traffic. ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 21
Further reasons • Influence driver route choice • Utilisation of reserve capacity on parallel arterials • Increased traffic safety (less congestion, safer merging) ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 22
When is ramp metering less helpful? Exit flow problems ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 23
Local Control Issues Note: ocr less sensitive than qcap (e. g. under adverse weather conditions) ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 24
Sample from Glasgow Implementation of ALINEA ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 25
Many additional tools • Real-time estimation of ocr • Real-time estimation of ramp queue • Ramp queue management • Various traffic-light policies • Switch on/off logic . . . ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 26
Why Coordinated Ramp Metering? No ramp queue constraints • Little efficiency improvement via coordination • Equity ? • Realistic ? • Diversion ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 27
Limited ramp storage capacity: • Full ramp Congestion “merely” retarded • Equity: bad significant improvement of both efficiency and equity via appropriate coordination ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 28
HERO: Ramp Metering coordination algorithm • Rule-based central control • employs (modified) ALINEA at each on-ramp • Master/Slave ramps for increased storage space • High efficiency (depending on available ramp storage space) ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 29
ALINEA Field Implementations >100 ramps in Europe (even w/o our involvement) Boulevard Périphérique, Paris, France • A 6, Ile-de-France • A 10, Amsterdam, Netherlands • Glasgow, Scotland • A 94, Munich, Germany • UK Highways Agency ramp metering roll-out • Tel Aviv, Israel • . . . Status: Mature tool, ready for immediate implementation ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 30
HERO/ALINEA Field Implementations • A 6, Ile-de-France (simplified) • Ile-de-France network (~ 80 ramps) • A 10, Amsterdam, Netherlands (~ 40 ramps) • Monash Freeway, Melbourne, Australia (6 65 ramps) Status: Mature tool, ready for immediate implementation ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 31
4. ROUTE INFORMATION AND GUIDANCE Multi-origin, multi-destination, multi-route per O-D pair. n Fixed direction signs: shortest path in absence of congestion n Rush hours n Changing demands, weather conditions, exceptional events, incidents underutilisation of infrastructure congestion, delays, reduced safety, increased fuel consumption, environmental pollution n ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 32
ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 33
VMS (Variable Message Signs) or two-way communication with equiped vehicles n Real-time information: – Drivers’ knowledge – Message length – Decision efficiency – System controllability – Travel time or queue length: drivers’ stress (e. g. BP in Paris) but also basis for route choice – Instantaneous (estimation) or predicted information n Route guidance – Control strategy ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 34
Our Concept: Simple feedback-based route guidance ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 35
Automatic Control of VMS in Aalborg, Denmark n Main goal: efficient crossing (northbound or southbound) of Limfjorden via BRIDGE (urban) or TUNNEL (motorway) n Particularly in presence of incidents, road works etc. n 116 loop detectors; 14 VMS in front of important bifurcations n Sample time: 1 min ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 36
n Two display modes (no real-time switch) – Delay information – Route Recommendation n Operational constraints: – Information is the most accurate available – Route recommendation according to user optimum principle – No incompatible VMS displays – Police report input ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 37
Aalborg network with VMS positions indicated. Bold black lines represent detector equipped segments. ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 38
VMS control modes: Delay information (a) and route guidance (b). ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 39
Automatic Control of VMS in the Interurban Scottish Highway Network n Motorway/Expressway Network n Geographically extended ( predictions necessary) but less complex in topology. n VMS: Combination of Information and Route Recommendation ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 40
n VMS Plans pre-approved (some 200) n Predictive/Feedback Strategy (Smith-Predictor type) n Off-line simulations n Implementation n Dummy Control n Evaluation/Comparison ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 41
The Scottish interurban network ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 42
ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 43
ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 44
VANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE
ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 46
n Glasgow integrated traffic control n Also involved: Boulevard Périphérique travel time information system Status: Mature for implementation ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 47
5. VARIABLE SPEED LIMITS ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 48
5. VARIABLE SPEED LIMITS n Increasingly popular control measure with many applications worldwide. n Several evaluations indicate substantial safety improvements (– 30 % accidents)… n … but no efficiency improvements (e. g. no reduction of travel times) n Why? ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 49
1) Very simple threshold-based control strategies An example of VSL switching logic. ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 50
The VSL switching logic at the Monash Freeway. ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 51
2) VSL impact on aggregate traffic flow not well known (a) n slope decrease n ocr increase n cross-points n capacity increase(? ) n capacity decrease (b) ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 52
3) Difficult fine-tuning of thresholds ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 53
New Tools 1) Automatic Fine-Tuning AFT 2 ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 54
AFT 2 simulated application for VSL control. ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 55
2) Different inputs: Slope of flow-occupancy curve 3) Feedback-based VSL upstream of bottlenecks as mainstream control ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 56
as ramp metering substitute ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 57
n No field-tested VSL control strategy available n Discussions with a road authority ongoing ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 58
6. URBAN SIGNAL CONTROL n Most available signal control strategies are less suitable under saturated traffic conditions Queue spillback – Wasting of green time – Increased delays (all movements) – Blocked exits – Accelerated queue increase – Queue spillback. . . – Serious infrastructure degradation. . . – Gridlock: Infrastructure breaks down ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 59
Traffic congestion = temporarily and locally d > smax + lack of efficient control ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 60
The TUC strategy n n n Based on modern control and optimisation methods Efficient under saturated traffic conditions as well Simple, easy to implement One measurement per link One communication with central computer per cycle. ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 61
TUC field-applications a) Chania, Greece (23 junctions) ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 62
b) Southampton, UK (53 junctions) ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 63
c) Munich, Germany (25 junctions) ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 64
n Independent evaluation report: The demonstrations proved that TUC is a robust and credible signal control strategy, both as a stand-alone system, as in Chania and Southampton, and as a hybrid system. Despite its limited fine tuning, TUC stood up very well against the well-established and well-fine-tuned resident systems of the test networks, which have very different characteristics with regard to network layout, detector locations within the links and traffic behaviour. ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 65
Other TUC Applications n Glasgow, Scotland n Macaé, Brazil (in final implementation phase) n Athens, Greece (planned) n Tel Aviv and Jerusalem, Israel (feasibility studies) Status: Mature tool; ready for immediate implementation ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 66
7. INTEGRATED URBANFREEWAY TRAFFIC CONTROL – Main control measures: Ramp metering n Signal control n VMS in metropolitan areas. n ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 67
– Drivers perceive the overall network as an entity – Aims: n Harmonise actions of various control measures n Avoid antagonistic actions n Maximize synergy – Several de facto implementations but very few methodological approaches: n Optimisation-based (LP) n Optimal control n LQ control: IN-TUC ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 68
Functional architecture of IN-TUC strategy ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 69
The Glasgow Site & the Goal ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 70
Effect of Control on Road Throughput and on Journey Times * Not statistically significant ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 71
8. MOTORWAY TRAFFIC SURVEILLANCE RENAISSANCE Travel-time prediction, queue-tail prediction etc. ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 72
Structure of RENAISSANCE ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 73
RENAISSANCE Applications n A 92, Munich, Germany (7 km) n A 3, Napoli, -Salerno (100 km) n Antwerp Status: Mature tool (estimation); ready for immediate implementation ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 74
9. OTHER CONTROL TOOLS Real-time toll plaza control ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 75
ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 76
Simulation Example: Work Zone M=3 μ = 1; = 2, 300 veh/h ADVANCED TRAFFIC CONTROL TECHNIQUES FOR FREEWAY SYSTEMS (AND URBAN NETWORKS), DSSL, TUC, GREECE 77