LO5: Simulation of transit signal priority strategies for brt operations
- 1. Simulation of Transit Signal Priority Strategies for BRT Operations Anna Matías Alemán January 15, 2013
- 2. Outline Motivation Background Objectives Methodology Evaluation Approach Boston Case Study Next Steps 3
- 3. Why TSP? A technique to provide priority to public transport services through traffic lights A means to improve performance of public transportation systems at the operational level A strategy aiming at reducing travel times and improving bus service reliability When properly designed for BRTs it can complement their other features such as exclusive bus lanes, off-vehicle fare collection, all-door boarding, etc., and potentially contribute to improved system performance 4
- 4. Types of Priority Unconditional – priority is always provided Conditional – priority is provided only when certain conditions are met (schedule adherence, vehicle load, etc.) TSP Actions With traditional signal controllers, the actions that can be used when a vehicle is detected are: Extension of the current green interval Early start of the green interval Skipping a phase Inserting an extra phase 5
- 5. Previous Studies in TSP The majority of studies in TSP are through simulation experiments It has been found that it can accomplish travel time reductions of 0-42% (Chang et al, 1995) But, as the buses benefit from this, cross-traffic and traffic light synchronization can be negatively affected Studies have shown that providing priority to all buses can significantly affect the overall traffic, therefore, providing priority to express routes -or select buses conditionally- could result in higher benefits (Dion and Rakha, 2005) Studies indicate that it may work well in arterial intersections in urban areas if properly planned and designed 6
- 6. Background Research Work to date has focused mainly on conventional bus systems; the study of TSP for BRT corridors is limited A recent simulation study of schedule-based TSP for a BRT line in Beijing Another study of TSP using speed guidance and advanced detection for a BRT line in Yingtan City But BRT corridors present a number of challenges and operating characteristics, which are different from conventional corridors and are worth considering (high frequencies of service, great levels of demand, etc.) 7
- 7. Research Objectives The objective of the proposed research activity is to study how TSP can benefit BRT systems and can best be incorporated into their operations both in the U.S. and in developing countries. The main tasks include the following: review literature and experience with TSP with both conventional and BRT systems, and document the main conclusions and lessons learned; determine how TSP can best be implemented in BRT corridors with different characteristics in terms of demand levels, frequency of service, etc.; evaluate TSP strategies that consider different conditions (schedule, headways, loading, etc.) in a BRT context; and develop guidelines for implementation of signal priority strategies in BRT corridors, based on their characteristics. 8
- 8. Data Sources Automated Passenger Counting Records (APC) Automatic Vehicle Location Records (AVL) Signal Timing Plans Traffic count studies/O-D Matrix from Planning Model Vehicle Specifications 9
- 9. Data Preparation Traffic Data Road network – coded or from a geographic file Traffic Flows – turning movements inputted from traffic counts or O/D matrix generated with Cube Analyst Signal Timing Plans 10
- 10. Data Preparation Transit Data Transit Routes delineated with defined route physical stop Average Departing headways and standard deviation calculated from AVL Records Average Arrival rates and Alighting Percentages per stop calculated from APC Records Vehicle type defined with capacity and dwell time parameters Average Initial load calculated for the segment of the route to be simulated, if necessary 11
- 11. Dwell Time Model Passengers are not really modeled in the simulation, but rather their effect on dwell times, which will be a function of the number of passenger boardings and alightings, using parameters that account for crowding and boarding and alighting times per passenger. Dwell time model: T=ɣ + αA + βB if there is no crowding T=ɣ + αA + βB’’ + (β +CF) B’’’ if there is crowding These parameters are defined by vehicle class to define the dead time component and the service time component of the dwell time. Therefore, the vehicle class will have a defined seating capacity, total capacity, dead time, alighting time, boarding time, and crowding factor. 12
- 12. Actions Evaluated Green Extension Red Truncation Phase Skipping (only if it is not a major cross-street) 13
- 13. Conditional Priorities Evaluated Schedule deviation (a bus must be running late) on all intersections -Headway limitation: 2 cycles & Lateness Treshold: 1min Combined schedule deviation and minimum passenger load on all intersections (a bus must be late and have the requisite number of passengers to be eligible for priority) Schedule deviation on all intersections with the minimum passenger load constraint only on critical intersections (cross- streets with high volumes or high-frequency bus routes) Schedule deviation on all intersections except on critical intersections (no priority will be provided in cross-streets with high volumes or high-frequency bus routes) *Will be evaluated in current conditions and other projections 14
- 14. Evaluation Metrics Impact on average and variability of bus travel times Impact on average bus speeds Impact on general traffic speed Headway variability Intersection delays Crowding Vehicle Emissions (in the long run) 15
- 15. The cities Case Studies Boston Silver Line 5 – Washington Street Limited bus-only lanes, 6 min headways during peaks, +15,000 weekday boardings Minneapolis Route 10 – Central Avenue Limited bus-only lanes, 12 min headways during peaks, +7,300 weekday boardings Santiago Routes 204 and 204 e – Carmen Avenue Limited bus-only lanes, 4 min headways during peaks ,+15,000 weekday boardings Many assumptions were made (fleet, frequencies, lanes, etc) 16
- 16. Boston • Silver Line 5 - Washington Street 17
- 17. Boston Case Study – SL5 Applied Conditional TSP (schedule-based) at “major” signalized intersections: Melnea Cass Boulevard Mass Ave – outbound East Berkeley St. Herald St. Operating Scheme: Bus computer sends location to MBTA’s Bus Control Center Bus Control Center checks if it is behind schedule and sends signal to hardware in kiosk on the side of the intersection Hardware in kiosk sends contact closure signal to the intersection signal controller which passes the signal to the BTD computer system BTD decides to grant priority (green extension or early green) 18
- 18. Boston Case Study – SL5 OTP Without TSP 19
- 19. Boston Case Study – SL5 OTP with TSP 20
- 20. Next Steps Finish gathering then inputs for the corridors Evaluate the effects of the different strategies on each corridor Project the scenarios to higher levels of demand and frequency (and visualize the effects of having more than one request per cycle) Generate conclusions depending on corridor characteristics 21