Simulation of Transit Signal Priority Using the NTCIP Architecture

Transit Signal Priority (TSP) is an important element of Bus Rapid Transit (BRT) that involves coordinated efforts between transit vehicle detection systems, traffic signal control systems, and communication technologies. Successful deployment of TSP requires thorough laboratory evaluation through simulation before field implementation. This paper presents the development and application of a simulation model specifically designed for the design and evaluation of TSP systems. The proposed simulation tool models in detail all the TSP components in accordance with the National Transportation Communications for ITS Protocol (NTCIP) standard for TSP systems. The study is intended to shed light on how the variety of TSP elements can be addressed in microscopic simulation in a structured and systematic fashion. Sample applications of the model on a real-life arterial corridor in California demonstrate its capabilities and features. Introduction Although past research and experience have demonstrated the benefits of Transit Signal Priority (TSP) to transit vehicles, skepticism still remains regarding its effectiveness among various parties. To address these skepticisms, evaluation Journal of Public Transportation, 2006 BRT Special Edition 118 methodologies that satisfy the concerns of a diverse set of stakeholders are needed (Gifford 2001). While field evaluation provides real world assessment, traffic simulation is advantageous in conducting “what if” studies before implementation and “before and after” analysis in evaluation. It is also a more economical way as compared to the cost of field evaluation. A TSP system is difficult to address in traffic simulation (Sunkari et al. 1995). Basic requirements for simulating TSP involve emulating the logic of fixed time/actuated traffic signals under the normal operation and during transit signal priority, detection of bus at the check-in and check-out points, priority generator, priority server, communication links between buses and traffic signals, bus movements in the traffic stream, and the dwell time at bus stops. Advanced features needed to be modeled include but are not limited to adaptive signal control, Automatic Vehicle Location (AVL) systems, additional priority treatment options (e.g., queue jump, transit phase, recall, green hold etc), on-line event monitor to record and report the status of buses and signals, bus arrival time predictor, on-line bus schedule checking, and passenger counting systems. TSP impact analysis relies greatly on simulation (Smith et al. 2005, Dale et al. 2000). Several commercial simulation software packages such as VISSIM (PTV 2003), CORSIM (FHWA, 2003), and PAMRAMICS (Quastone 2004) provide, to some extent, functions for simulating traffic signals and transit vehicles. Evaluation of TSP has been conducted mainly through these simulation tools. Recent examples of this include the work of Balke et al. (2000), Davol (2002), Shalaby et al. (2003), Dion et al. (2004), and Ngan et al. (2004) who used CORSIM, PARAMICS, and VISSIM to evaluate the effectiveness of the early green and the extended green strategy. Most simulation models currently available lack most of the characteristics and capabilities for realistically modeling real-life TSP systems. Application of oversimplified simulation models may draw inconvincible conclusions and sometimes mislead the implementation. In addition, the extensive use of AVL data in transit management, planning, and operation has presented a challenge to the development and application of next generation traffic simulation tools (Chu et al. 2004). A new NTCIP standard (NEMA/ITE/AASHTO 2005) is being developed that aims to define communication protocols and the logical architecture of a transit signal priority system. It is extremely important for the design of future TSP simulation models to comply with the NTCIP definitions so that the diversity of the transit signal priority systems can be addressed in a systematic manner. Simulation of Transit Signal Priority Using the NTCIP Architecture 119 This paper presents the development and application of a simulation model specifically designed for the design and evaluation of various TSP systems. The proposed simulation tool models virtually all the TSP components in accordance with the NTCIP definitions. The model was developed in support of a study for developing advanced bus signal priority strategies sponsored by the California Department of Transportation (Caltrans) in cooperation with the San Mateo Transit District (SamTrans) in the San Francisco Bay Area. Logical and Physical Structure of a TSP System Logical Structure of TSP NTCIP provides both communication protocols and the vocabulary (called objects) necessary to allow electronic traffic control equipment from different manufacturers to operate with each other as a system. Two main NTCIP standards that are related to traffic signal control and transit signal priority control are NTCIP 1202 and NTCIP 1211. The former defines the commands, responses and information necessary for the management and control of actuated traffic signal controllers. The NTCIP 1211 Signal Control Priority standard provides the framework and communication protocols for the design of a signal priority system. One of the significant contributions of NTCIP 1211, aside from the description of the “computer objects” for communication, is the definition of the functional entities of a TSP system. As shown in Figure 1, the logical structure of a TSP system is composed of a Priority Request Generator (PRG), a Priority Request Server (PRS), and a Coordinator. The primary functions of the PRG are to determine the Figure 1. Logical Structure of a TSP System (NTCIP 1211) Journal of Public Transportation, 2006 BRT Special Edition 120 necessity for generating a priority request, to estimate priority service time, and to communicate the request to the PRS. The final decision is made in the PRS. It receives priority requests from multiple PRGs, processes the requests based on importance and priority, and sends the selected requests to the traffic signal controller for priority operation. Physical Structure of TSP According to the Intelligent Transportation Society of America (ITSA) (2003), a physical TSP system is composed of three major components: the vehicle detection system that detects transit vehicles and generates priority requests, the traffic signal control system that receives and processes the request for priority at the intersections, and the communications system that links the vehicle detection system with the traffic signal control system. The bus detection system is further categorized into point detection or selective vehicle detection (SVD), zone detection, and area detection systems. As illustrated in Figure 2, inductive loops and radio frequency (RF) tags with readers are two typical point detection devices. The vehicle-to-controller communication is achieved through on-vehicle equipment (either a transponder or a RF tag) and a road-side receiver. The on-vehicle device contains a data packet that is sent to the receiver when a bus passes through the detection point. Upon the detection of a transit vehicle, signal priority can be operated at the local or the central level. Unlike point detection devices that sense the presence of transit vehicles at fixed locations, zone and area detectors may extend the detection area to a certain distance from the intersection. The OpticomTM system from 3M is probably the most widely implemented traffic signal priority control system that enables signal priority operation to both emergency and transit vehicles. As shown in Figure 3(a), the system works by an emitter mounted on the vehicle. When activated, it sends an optical flashing signal at a certain rate and at an exact duration, emergency vehicles and buses are differentiated by different flashing frequencies. Figure 3(b) depicts an AVL based system, in which a bus provides schedule adherence and passenger information along with the priority request continuously to the traffic/ transit management center, where the central computer in return makes decision upon whether and how the transit vehicle should be served. A major distinction between the zone/area detection based and the point detection/SVD-based TSP lies in the control logic with regard to the initiation time of the priority operation. SVD-based systems initiate the priority operation upon Simulation of Transit Signal Priority Using the NTCIP Architecture 121 Fi gu re 2 . Po in t D et ec ti on B as ed T SP S ys te m s Journal of Public Transportation, 2006 BRT Special Edition 122 Fi gu re 3 . Zo ne a nd A re a D et ec ti on B as ed T SP S ys te m s