2020

Farzana Chowdhury, University of Texas, Arlington
Peirong (Slade) Wang, University of Texas, Arlington
Pengfei (Taylor) Li, University of Texas, Arlington

Traffic signal priority (TSP) is a common operation at signalized intersections to accommodate the green requests from special vehicles like buses or ambulance as well as emerging connected automated vehicles (CV). While most TSP strategies today allow for placing the TSP requests only 20 to 30 seconds away, special vehicles or CVs can transmit TSP requests via mobile computing technique minutes before they arrive at the downstream intersections. A unified optimization framework for traffic signal priority scheduling based on the proposed phase-time network (phase-time network) models is presented. Multimodal traffic signal priority requests are defined as the requests from various types of vehicles (e.g. fire trucks, ambulance, buses, CAVs). Two mathematical programming formulations are presented: (i) Mixed Integer Linear Programming (MILP) formulation and resilient MILP formulation (R-MILP). The R-MILP formulation is modified from the MILP formulation in a way such that MIP solvers can always reach a mathematically feasible solution. Based on the R-MILP results, a policy is developed to accept or reject excessive TSP requests according to their priorities and time windows. In order to schedule the TSP requests considering the background traffic, a search algorithm is also developed in the modified phase-time network. In the modified phase-time network, each arc cost is dynamically estimated according to traffic queuing states before and after the corresponding traffic signal operations.  This research aims to provide on-line optimization engines for multimodal TSP scheduling with and without considering the background traffic at signalized intersections, especially when the TP requests are excessive.

Hyeonmi Kim, Korea Transport Institute (KOTI)
Yao Cheng, University of Maryland, College Park
Gang-Len Chang, University of Maryland, College Park

Despite the increasing deployment of bus dedicated lanes in practice, commonly available programs for arterial signal plans are not designed to provide concurrent progression for such bus flows and general traffic. Hence, this study presents a bandwidth maximization model for traffic engineers to design both phase sequences and offsets over a series of intersections along an arterial with bus dedicated lanes. The proposed model, offering concurrent progression to both modes based on traffic volumes and geometric conditions, has taken into account the following factors that may affect the progression effectiveness: bus dwelling time at bus stops, location of bus dedicated lanes, traffic queues, and conflicts between through bus movements and left-turning traffic. In addition, by weighting the bandwidths with the passenger volumes by mode and by direction, the proposed model is capable of offering progression to the mode(s) and the direction(s) for which it is justified from the perspective of maximizing the benefits to all arterial users. The results from numerical analysis with two arterial sites with planned BRT have confirmed the effectiveness of the proposed model in producing concurrent progression bands for both modes. Evaluation with extensive simulation experiments and the MOEs of delays and the number of stops has demonstrated that the benefits from concurrent progression by the proposed model will not be at the cost of other MOEs. Hence, given reliable information for all key input variables, the proposed model can serve as one effective tool for traffic professionals to do the essential signal design task.

Awad Abdelhalim, Massachusetts Institute of Technology
Montasir Abbas, Virginia Polytechnic Institute

This study assesses the serviceability of a cooperative Transit Signal Priority (TSP) strategy utilizing vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication in a Connected and Autonomous Vehicles (CAV) environment.  In a prior work, we proposed the Connected Bus-Holding Transit Signal Priority (C-BHTSP), which was implemented for assessment in VISSIM using an advanced vehicle actuated control logic to emulate transit vehicles' V2V and V2I communications for a detailed isolated intersection using high quality data including up-to-date vehicle flows, signal timing, transit schedules, actual transit arrival, and departure and dwell times. Our proposed strategy significantly improved transit performance while causing minimal impact to non-transit traffic compared to conventional TSP. This study aims to further assess the feasibility and serviceability of C-BHTSP in a CAV environment under varying CAV market penetration rate (MPR) and different scenarios of signal controller to CAV information exchange. The results of the study conclude that a CAV MPR of only 20% was sufficient to overcome the negative impact of implementing the C-BHTSP strategy on non-transit main arterial traffic while maintaining transit benefits, and higher MPRs produced substantial improvements in travel time, delay, fuel consumption and greenhouse gas emissions.

SMA Bin Al Islam, North Carolina State University
Ali Hajbabaie, North Carolina State University

This paper presents a Cell Transmission Model (CTM) based formulation for multi-modal traffic signal timing optimization in urban street networks. Existing multi-modal CTM based formulations (used in signal timing or traffic assignment) involve the conversion of transit flow to equivalent passenger car flow thus, cannot represent the multi-modal flow dynamics with reasonable accuracy when the movement of transit vehicles are prioritized. In such conditions, existing CTM models transfer transit vehicles from one cell to the next one, before the vehicles in front of them violating the first-in-first-out queuing discipline, which is different than what happens in real-world. The proposed optimization program introduces constraints to update the position of transit vehicles within the CTM, based on the speed and capacity variations between different classes of vehicles. The problem is formulated as a mixed-integer nonlinear problem; thus, it is implemented in a distributed model predictive control framework to find solutions in real-time. The proposed formulation is executed in a simulated arterial street with four signalized intersections in Springfield, IL using different traffic volume levels and transit vehicle frequencies under various market penetration rates of connected vehicles (CV). The results showed that the proposed algorithm performs better than the conventional transit signal priority-based (CTSP) systems under all tested demand patterns and bus frequencies. At 0% CV penetration rate, the proposed formulation reduced the average bus delay by 1% to 70% and car delay by 52% to 76% compared to CTSP. At 100% CV market share, 49% to 80% further reduction in average bus delay was observed compared to the 0% CV penetration rate under different demand and bus frequencies tested.

Qinzheng Wang, Leidos, Inc.
Xianfeng Yang, University of Maryland, College Park
Blaine Leonard, Utah Department of Transportation
Jamie Mackey, Federal Highway Administration (FHWA)

In 2017, a connected vehicle (CV) corridor utilizing dedicated short-range communication (DSRC) technology was built along the Redwood Road in Salt Lake City, Utah. One main goal of this CV corridor is to implement transit signal priority (TSP) when the buses are behind the published schedule by a certain threshold. With the data generated by the transit vehicles, transmitted through the DSRC system, logged by traffic signal controller, and coupled with the UTA data from transit operation systems, some performance of TSP can be analyzed including TSP requested, TSP served, bus reliability, bus travel time, and bus operation time. For providing better signal progression to both buses and passenger cars, the signal plan for this CV corridor was updated in October 2018 and this research aims to compare the TSP performance before and after the signal retiming. Field data of August, September, November, and December of 2018 are selected to conduct a comprehensive analysis. Results show that the TSP served rate after signal retiming is 35.29% which is higher than that of 33.12% before signal retiming. Also, compared with the cases before October, the bus reliability along the northbound and southbound of the CV corridor was improved by 2.4% and 1.47%, respectively, and the bus travel time and bus operation time were reduced as well.

Murat Bayrak, Pennsylvania State University
S. Ilgin Guler, Pennsylvania State University

Transit signal priority (TSP) can be used to improve bus operations at intersections. However, implementing TSP can often increase the delay of non-transit modes. Hence, it is necessary to evaluate the impacts of TSP both on car and bus operations to determine optimal locations to equip with TSP to improve network operations. To do so, the link transmission model is used to evaluate the travel times of both cars and buses on the network while accounting for dynamic queuing and queue spillover. This method is then used to evaluate different combinations of locations for TSP implementation and determine the optimal configuration that can minimize the total travel time of network users, including bus and car passengers. Sensitivity of the proposed algorithm to demand level, changes in transit network, implementation strategy and solution method are also evaluated. For all tested scenarios, the found optimum TSP configurations achieve significant reduction of total bus passenger travel time while creating minimal impact on total car travel time. The results reveal that in general, not all intersections should be equipped with TSP and central intersections within a network are promising locations for TSP implementation to reduce total travel time of network users. Additionally, it is found that the total travel time of network users can be further decreased by only activating TSP for buses with more than a certain number of on-board passengers.

A. M. Tahsin Emtenan, CATT Laboratory
Christopher Day, Iowa State University

In recent years, Automated Traffic Signal Performance Measures (ATSPMs) have emerged as a means of developing situational awareness of traffic conditions at intersections and for assessing the quality of signal operations. As a growing number of agencies are adopting the technology, there is a need to understand how detector configurations can influence the outcomes of an analysis using ATSPM. Current practices on detector configuration vary considerably from one agency to another; at one extreme, agencies may use one single detector input channel per phase without regard to where those detectors are located, where at the other extreme, some agencies may utilize all possible channels to observe each individual lane at multiple positions. There are also variations in the design of detection zones (lengths and positions). This study takes on the problem in two parts. The first of these examines the impact of stop bar detection zone length and lane- or approach-based detector assignment on the ability of performance measures to accurately identify whether split failures occur. The second part examines the impact of setback detector distance on the use of a “percent on green” metric to serve as a proxy measurement of the number of stops. The paper presents recommendations for performance measure calibrations and detector configurations that follow from these outcomes.

Pruthvi Manjunatha, University of Florida
Xi Duan, University of Florida
Marian Ankomah, Fekams Company Limited
Lily Elfeteriadou, University of Florida
Sivaramakrishnan Srinivasan, University of Florida
Raj Ponnaluri, Florida Department of Transportation

The Florida Department of Transportation (FDOT) has proposed the implementation of Adaptive Signal Control Technologies (ASCT) on eight corridors in Florida to overcome the limitations of traditional signal systems in cases of changes in traffic demand, weather, incidents, etc. The objective of this paper is to examine the relationship between site characteristics and the effectiveness of ASCTs using performance measures obtained from field data. First, the mobility and safety benefits of the ASCT implementation are assessed by comparing performance measures of time of the day (TOD) plans versus ASCT through field data collection. Two critical intersections are identified within each corridor and performance measures such as corridor travel time, intersection delay, major and minor street queues, turning movement etc. are collected. Crash data are collected over a period of fifty-nine months for safety analysis. The summary field data are used to build regression models of performance measures as functions of site characteristics. Qualitative observations and institutional issues are obtained by interviewing local staff. Recommendations are made on the suitability of corridors for ASCT implementation, and guidelines are provided for effective field implementation.

Nemanja Dobrota, Kittelson & Associates, Inc.
Aleksandar Stevanovic, University of Pittsburgh
Nikola Mitrovic, CHA Consulting, Inc.

Adaptive Traffic Control Systems (ATCSs), have been in use since the late ‘70s but have only received significant attention in the U.S. during the last 10 years. Since 2008, the number of these systems has increased by more than 600%. A brief literature review clearly shows a shortage of comprehensive studies that gather data from numerous field evaluations, categorize those evaluations based on a number of factors, summarize findings, and develop meaningful conclusions. This study addresses this gap by presenting a framework and a tool which enable comprehensive analysis of ATCS’s deployed and evaluated in the U.S. Compared to some previous studies, where the focus was on brief and general experiences with ATCSs, tools presented in this study allow detailed analysis of ATCS deployments and investigation of numerous criteria important for their evaluations. Relevant ATCS data from deploying agencies and literature are collected, discussed, and used to populate a database of the Assessment Tool for Adaptive Traffic Control ((AT)2C). The main purpose of the (AT)2C is to help practitioners and researchers to identify, compare, and  assess statistics of relevant ATCS technologies, mainly from the perspective of their field benefits.

Ali Danesh, Tongji University
Wanjing Ma, Tongji University
Ling Wang, Tongji University
Xiaolong Ma, Qingdao Hisense Transteh Co., LTD
Ali Gholami, Golestan University

In urban road network, intersections are the main bottlenecks. Selecting an appropriate intersection control type can significantly improve capacity, delay, and safety. Several studies have proposed guidelines in selecting the appropriate control type. Nevertheless, their recommendations were mostly made based on achieving the lowest delay, and they did not propose detailed guidelines by considering capacity as measure of effectiveness. Additionally, they did not include signalized roundabout into comparison. Therefore, this paper aims to propose recommendations to select the most efficient control type among two-way stop (TWSC), signalized intersection, unsignalized roundabout (RB), and signalized roundabout (SIGRB) for an isolated intersection. The measures of effectiveness used for the comparison are capacity and delay, and the procedure to calculate these two measures is taken from the Highway Capacity Manual 2010. The sensitivity analyses with regard to the impact of flow rates and turning ratios on the measures were evaluated. The results showed that in the case of low to medium flow rates, TWSC gives the highest capacity and RB provides the lowest delay. For high flow rates, however, SIGRB outperforms other control types both in capacity and delay. It was also observed that the increase of turning ratio can deteriorate the delay and capacity of all control types while its impacts on SIGRB were the least. The findings demonstrated that a roundabout should be put into consideration if an intersection improvement is needed. The result of this paper can help decision-makers to decide the best control type for an isolated intersection.