Integrated Pedestrian Simulation in VISSIM PTV worldwide MUGS Conference October 30-31, 2008 Wellington Founded > 1979 30 years fast approaching Employees > Approximately 800 persons in the PTV Group worldwide Offices > Europe > North America > Middle East > China > Asia-Pacific PTV Group Partner www.ptvap.com PTV AP 2008 2 Key market positions Business units Traffic Planning and Optimisation > Global #1 with PTV Vision suite (VISUM, VISSIM) Traffic Data Management > Innovative Management Systems Mobility & Navigation Solutions > Global leader and pioneer Advanced telematics components (PAYD, Toll-Collect) > World leader Logistics Planning (software and consulting) > European leader Pedestrian Modelling > Scientificaly developed, comprehensive, integrated solution PTV Asia-Pacific Traffic Mobility Logistics www.ptvap.com PTV AP 2008 3 www.ptvap.com PTV AP 2008 4 The Software Suite for the Transportation Professional Organisations using PTV Vision Software Data Quality Check VISSIM Simulation Modelling VISUM Strategic Network Modelling Real Time Information www.ptvap.com PTV AP 2008 5 www.ptvap.com PTV AP 2008 6
www.ptvap.com PTV AP 2008 7 PTV Asia-Pacific International Users - PTV Vision > Regional business unit of PTV worldwide > Permanent offices in Asia-Pacific since December 2005 > Singapore, Brisbane & Melbourne > PTV Software Key Users in Asia-Pacific > Queensland Transport, TransLink > NSW RTA Network Planning > VicRoads Design and Metropolitan regional offices > Land Transport Authority, Singapore > Hong Kong Airport Authority > Hong Kong Mass Transit Rail Corporation > Local authorities, consultants, universities > Thirty staff and growing www.ptvap.com PTV AP 2008 8 Normal infrastructure operation > Capacity analysis > Level of Service > Queuing > Travel Times > Rail Platforms > Transit Stations > Public Transport Interchanges > Airports > Town Centres www.ptvap.com PTV AP 2008 Normal infrastructure operation > Effect on Traffic Flow (Capacity Constraint) > Pedestrian Crossings Urban Planning > Design Conceptualisation > Visualisation > Shared Space Applications Exhibition Road, London, UK Brandenburg Gate, Berlin, Germany www.ptvap.com PTV AP 2008 www.ptvap.com PTV AP 2008
www.ptvap.com PTV AP 2008 Evacuation > Safety Assessment of Building Design > Retail (Shopping Centres) > Commercial (Offices) > Recreational (Hotels) > Sports Stadiums > Concert Arenas > Large Scale Carnival Events The Research Pedestrians in VISSIM The Social Force Model > Pedestrians have always been in VISSIM as an important aspect in junction modelling > A simple link based behavior model was sufficient for these applications > pedestrian flows play a more important role in other situations: > bus terminals > railway and underground stations > Airports > buidlings (retail, residential, commercial) > sports stadiums > Developed by Prof. Helbing, Federal Institute of Technology Zurich, 1995 > Each pedestrian is influenced by a number of forces, > repulsive forces from other pedestrians, > repulsive forces from borders and > driving force towards the desired direction of motion. Forces from fellow pedestrians Forces from borders > For these applications, a more sophisticated area based behaviour model was included in VISSIM : The social force model Driving force in the desired direction of motion trail formation www.ptvap.com PTV AP 2008 15 www.ptvap.com PTV AP 2008 16 The Social Force Model some details Model calibration: empirical data > In the Social Force Model, the acceleration (change of velocity) of pedestrian α is composed from a deterministic and a stochastic part. > Measurement technique: video analysis > The deterministic part f is the sum of an acceleration toward the preferred velocity, repulsive forces from boundaries, repulsive forces from other pedestrians, attractive forces from other pedestrians. www.ptvap.com PTV AP 2008 17 www.ptvap.com PTV AP 2008 18
www.ptvap.com PTV AP 2008 19 Conditions of pedestrian characteristics Model calibration: Experiments > Results of measurements of pedestrian traffic depend on many variables. Source: Technical University of Delft Source: Technical University of Dresden > Temperature > Gradient age Hour of Day Source: T. Kretz, University of Duisburg- Essen www.ptvap.com PTV AP 2008 20 Flow through a bottleneck: dependence on the width VISSIM Pedestrian Model Development www.ptvap.com PTV AP 2008 21 Extensions to VISSIM simulation mechanics How pedestrians find their way: potential fields > movement of objects in areas, not lane-oriented > Rectangles or polygons > The VISSIM network infrastructure is transformed to general 2D areas. Potential fields are created so the pedestrians can navigate through the most complex environment. > interaction of flows in different directions on the same lane VISSIM network Potential field for navigation www.ptvap.com PTV AP 2008 23 www.ptvap.com PTV AP 2008 24
www.ptvap.com PTV AP 2008 25 Modeling pedestrian facilities in VISSIM > sources defining multiple flows and obstacles routes destinations from by sub-areas as different sub-areas of pedestrians sources as intermediate and destinations destinations Pedestrian attributes > user defined pedestrian types > physical dimensions > social force parameters > desired speed distribution www.ptvap.com PTV AP 2008 26 Modeling principle: everything is defined by areas > pedestrians move on areas > special areas for entries, exits, counters, routing etc. Model calibration: counterflow simulation > many other pedestrian modelling packages produce full blockages > in reality, pedestrians organise themselves into lanes Entrance Entrance Desk Desk Exit Desk Exit Restroom Restroom www.ptvap.com PTV AP 2008 27 www.ptvap.com PTV AP 2008 Evaluations Project Methodology > Travel Time: > from any area to any area > Locally > density, > average speed and > flow in an area > Level-of-Service > density based FRUIN-LOS (a common measure in pedestrian design) > Time-aggregated > occupancy of areas > Similar to traffic studies!! > Scope > Base Year > Data Collection > Composition > Site Visits > Model Build > Calibrate > Performance Measures > Scenarios > Data Collection video observation / CCTV assists www.ptvap.com PTV AP 2008 29 www.ptvap.com PTV AP 2008 30
Case Study: North Melbourne Station Services Demand North Melbourne North Melbourne Flagstaff Melbourne Central Flagstaff Melbourne Central Southern Cross Flinders St Parliament Southern Cross Flinders St Parliament Growth Areas Aerial View PTV worldwide Concourse View (from PTV north) worldwide Melbourne CBD www.ptvap.com PTV AP 2008 35 www.ptvap.com PTV AP 2008 36
www.ptvap.com PTV AP 2008 37 Base Model Development > Morning Peak hour model (7:30-8:30) > Platforms / Concourse represented as Pedestrian Areas > Furniture/ Shelter represented as Obstacles (physical barrier) > Platform and Concourse levels connected with Ramps > Concourse currently only at northern end of station > Six Platforms > Three Southbound (Up Direction) > Three Northbound (Down Direction) > Various Rolling Stock (Regional/ Metropolitan services) > Number of Doors > Number of Services per Platform > Boarding and Alighting Passenger Numbers > Passenger Interchange Patterns per Service Base Model Development > Dwell Time at station based on calculation: occupancy + boarding + alighting > Five classes of stationary time > Boarding volumes based on spare capacity > If train is near full, not all passengers can board > Passengers alight before passengers board > Site Investigations identify that: > Crowding on platform near to shelters > High volumes of passengers alight near to ramps (placement on train to minimise journey times) www.ptvap.com PTV AP 2008 38 PTV worldwide Future Scenario Options managed within spreadsheet Change of Rolling Stock (change capacity) Additional Services (add more rows) Impacts of Service Delays (change in arrival times) Change Passenger Volumes or Interchange Distribution Change in Station Operations (change in platform arrival) www.ptvap.com PTV AP 2008 39 www.ptvap.com PTV AP 2008 40 Integrated Pedestrian Simulation in VISSIM MUGS Conference October 30-31, 2008 Wellington