Increasing ridership on transit systems has long been recognized as one of the potentially effective strategies for mitigating urban traffic congestion from the demand side. However, the complex nature of urban traffic congestion and the inevitable frictions between transit vehicles and passenger cars often negate the effectiveness of operational strategies for enhancing the service quality of a transit system. Hence, it is essential to have an effective control platform that can properly integrate available strategies in both offline and real-time operations. In need of such a platform, this study proposes an integrated operating system that consists of a traffic monitoring module, pre-timed signal optimization models, and real-time control strategies.

In the proposed system, the traffic monitoring module is to systematically convert the real-time bus GPS data into geographic-based sensors that can estimate both bus and general traffic conditions in urban networks beyond its original function of monitoring each buses speed and locations. Taking advantage of heavy bus flows and spatial evolution over a link, the proposed pseudo detection module can provide more extensive information (such as link and movement-specific speeds) than conventional point-measurement detectors. The system can also provide the spatial and temporal variation of general traffic conditions by converting the time-varying bus speed in each sub link, viewed as a pseudo sensor.

The second part of the system constructs a bus-based signal coordination model to optimize the offsets for an urban arterial that needs to accommodate a high volume of bus flows. The corridor signal model for bus-based progression considers average bus dwell time, relatively slow bus speeds, physical constraints of bus stops, queue formation/discharging process and its relationships with the bus stop location. The enhanced model (BUSBAND) is designed to optimize the offsets at each intersection with the objective of maximizing the total effective bus bandwidth which accounts for the dwell time uncertainty.

Due to the stochastic nature of bus traveling and passenger arriving patterns, the pre-timed control may fail to contend with the fluctuation of bus arrivals at the intersection. Hence, this study has further developed a real-time signal control system which integrates an active TSP with the pre-timed BUSBAND. The integrated system dynamically adjusts the green time (i.e., early green or green extension) based on the real-time detected bus locations to facilitate the bus progression and expand bus bands. In order to best the benefits of the integrated TSP operations, the proposed system also has a module to select an intersection for deploying an active TSP, based on the pretimed signal plan.