Traffic Flow Theory and its Applications in Urban Environments An Innovative Approach Presented by Dr. Jin Cao 30.01.18 1
Traffic issues in urban environments Pedestrian 30.01.18 Safety Environment 2
Traffic interacts with other systems HIGHWAY URBAN Urban Transport & Mobility Land Use Infrastructure Quality of life Energy Technology Environment 30.01.18 3
An innovative approach? Use Traffic Flow Theory in the Macroscopic Modelling of the Dynamics and Interactions of Urban Systems Urban traffic and parking systems 30.01.18 4
Dynamics and interactions between two urban systems traffic and parking Traffic control Traffic Why congested? Cruising-for-parking Land use More parking spaces? Induced demand Parking supply 30.01.18 5
Global issue: cruising-for-parking Status quo - studies City 1977 Freiburg 1985 Cambridge 1993 New York 2005 Average (16 cities) The High cost of Free Parking 2011 Average (20 cities) IBM Global Parking Survey Cruising time Traffic share Traditional data collection 12 min 30% 8 min 8% Cruising distance Non-transferrable conclusions 6 min 74% (to other cities) Trip data is hard to acquire Is there a generic approach to estimate cruising conditions? 8 min 30% 5794 km/day (L.A.) 20 min 30.01.18 6
Global issue: cruising-for-parking Status quo - solutions Smart parking (data) Real time parking data: cannot remove bottleneck cannot estimate cruising traffic cannot solve competition Parking policy GPS data to estimate cruising? Under development Only answers HOW (behavior) but not WHY What is the mathematical relation between parking and traffic? Performance-based pricing: cannot fully remove cruising slow reaction cannot reflect system dynamics 30.01.18 7
Urban traffic difficult to model Traditional traffic flow theory not always applicable Single roads Challenges Network Trajectories Time Car-following models Fundamental diagram Queuing theory Trip data collection y Time x 30.01.18 8
Traffic + Parking more complex Parking causes cruising Parking could lead to congestion Congestion affects parking Congestion hinders drivers to park Challenges Model the dynamics of both systems Incorporate the dynamics to calculate the probability of finding parking 30.01.18 9
Vehicular flow on urban networks Individual travelers Leave the area Depart parking Find parking Start to search Transition steps Enter the area 30.01.18 10
Traditional method Collect travel time of each individual traveler Trips made Individual steps Plot trajectories Time step diagram Aggregate info Cumulative plot Leave the area Steps 1 2 Cumulative N 2 Depart parking Find parking Start to search 1 Enter the area 1 2 time time 30.01.18 11
Macroscopic Traditional method Collect travel time of each individual traveler Trips made Individual steps Plot trajectories Time step diagram Aggregate info Cumulative plot Leave the area Steps 1 2 Cumulative N 2 Depart parking Find parking Start to search 1 Enter the area 1 2 time time 30.01.18 12
Macroscopic method Estimates car exchange between different states Trips made Individual steps Aggregate info Cumulative plot Leave the area Cumulative N Enter the area Start to search Find parking Depart parking Find parking Start to search Depart parking Leave the area Enter the area Time Slices Time 30.01.18 13
Macroscopic method Provides macroscopic outputs regarding cruising Trips made Individual steps Aggregate info Cumulative plot Leave the area Cumulative N searchers Enter the area Start to search Find parking Depart parking Find parking Start to search Enter the area Total travel time Total cruising time Depart parking Leave the area Time Slices Time 30.01.18 14
Macroscopic method Evaluates system dynamics of cruising conditions Parking system Number of users Parking occupancy (%) Cruising system Number of cruising vehicle Next time slice Probability of finding parking Traffic system Number of users Traffic density Travel speed 30.01.18 15
Macroscopic method - formulation Parking condition Traffic condition This image cannot currently be displayed. based on Macroscopic Fundamental Diagram 30.01.18 16
Macroscopic method - formulation Searched time (min) Contour plot of the average probability of finding parking Number of free spaces in relation to the number of searchers 30.01.18 17
Macroscopic method - formulation Vehicles go through each transition Vehicles in each relevant state 30.01.18 18
Validation of the model (Zürich) Case study Model Inputs Rennweg 2687 trips Traffic arrival to the network (MATSim) Parking supply Road network r = 300 m 0.28 km 2 Flow (veh/h lane) Density per lane (veh/km) Macroscopic fundamental diagram of Zürich (SVT) Parking durations (MATSim) 30.01.18 19
Validation of the model (Zürich) Parking occupancy Parking occupancy 100% 95% 90% 85% 80% 75% 70% 65% 60% 11.30-15.30 <5 spaces available real data transformed model results R-squared=0.97 9 10 11 12 13 14 15 16 17 18 19 Hour of the day Cruising at 12.00: 30 cars searching 2 spaces available 13 min cruising time 60% of share No speed drops Average cruising time before parking (minutes) Average cruising time Hour of the day In one day: 83 hrs additional travel time 1038 km cruising distance 30.01.18 20
Contributions and applications Academic Mathematical relation between parking availability & traffic condition Multiple systems Governing Policy-wise Practical Parking supply Time control Pricing Enforcement On-/off-street Generic Dynamic Technology -wise Evaluate the effectiveness of new technology Macroscopic Low data requirements and computational effort Information -wise Forecast Large event management Modal choice based on full travel time 30.01.18 21
Application parking policy Reduced supply Scientific foundation for Dynamic pricing Cruising time Constant supply Parking demand Current supply Hour of the day Increased supply Policy tests - quantify effects on traffic Hour of the day 30.01.18 22
Application IPS evaluation (Intelligent Parking System) 10:30-11:30 Reduction 59% 15:00-16:00 Reduction 51% Other time Reduction 7% Hour of the day 30.01.18 23
Further development of the model Investigate the distribution of parking and traffic in urban networks Incorporate big data and machine learning into the model to improve the accuracy Evaluate the impact of connected / autonomous vehicles on traffic and parking 30.01.18 24
Future research Urban Traffic Management Multimodal transport system Infrastructure Emergent technologies Land Use Infrastructure Quality of life Energy Technology Environment 30.01.18 25
Thank you for your attention! Presented by Dr. Jin Cao Reference Cao, J. and Menendez, M., 2015. System dynamics of urban traffic based on its parking-related-states. Transportation Research Part B: Methodological, 81, pp.718-736. Cao, J., Menendez, M. and Waraich, R., 2017. Impacts of the urban parking system on cruising traffic and policy development: the case of Zurich downtown area, Switzerland. Transportation, pp.1-26. 30.01.18 26