Despite the significant progress made by existing studies on optimizing arterial signal design, the impacts of multiple high-volume path flows and the mutual impedance between transit vehicles and general traffic on the progression effectiveness have yet to be addressed. A congested arterial traffic system, designed to maximize the bandwidth for through traffic flows, is likely to be plagued by frequent turning queue spillbacks and the inevitable mutual blockage between the through and turning traffic if the traffic patterns comprise multiple high-volume path flows. The lack of accounting for the excessive delay caused by the interference between transit and general traffic flows in the existing progression-maximizing methods also renders their applications to arterials of heavy mixed traffic flows ineffective.

In response to the challenge of accommodating multi-path traffic patterns and heavy-mixed flows in arterial signal design, this study has focused on developing a real-time adaptive multi-path signal control system that can dynamically generate the most effective control strategies to contend with potential turning-bay overflows, the mutual blockage between critical path-flows, the mutual blockage between general traffic and transit flows. The developed real-time control system is designed with a customized structure to proactively respond to time-varying traffic demand with adaptive yet sufficiently stable signal plans at either the intersection or the arterial level.

The first primary model embedded in the developed real-time control system is designed to offer a multi-path progression signal plan for the arterial to minimize the impacts of any related factors that may degrade its overall progression effectiveness, including excessive turning queues and the mutual blockage between different major path flows.The developed real-time control system’s second primary model is proposed to incorporate the control objective of bus delay minimization in the design of maximizing progression for general traffic. The bus-friendly signal progression plans generated from this model can account for the unique operational characteristics of transit vehicles and minimize the mutual impedance between the passenger car queues and dwelling bus flows on each link.

To contend with the variability of traffic volumes and changes in path-flow patterns over different time periods, the study has further integrated these two offline models along with essential supplemental modules to construct a real-time adaptive control system. With a customized hierarchical control structure, the developed real-time system will allow the users to generate the optimal signal plans based on multiple preferred MOEs, and also offer the time window for adaptively changing the high-level produced signal offsets at the local level based on the most updated information.

The developed systems, evaluated extensively under simulated offline and real-time experimental scenarios, has demonstrated their effectiveness in achieving the purported control objectives, and can serve as a reliable tool for the design of signal progression plan, especially for major arterials accommodating mixed traffic, experiencing heavy volumes over multiple traffic paths, and constrained by insufficient turning bay length.