Objective-Driven Traffic Signal Actuation using Vehicle Trajectory Data: Dilemma Zone Protection,Split Failure Mitigation,and Dynamic Coordination

Abstract

With connected vehicle technology and enhanced sensors coming to market, vehicle trajectory data is becoming increasingly available for signal control applications. Few previous studies have explored the use of this data for actuated control. This paper explores applications of real time vehicle trajectories at high penetration rates expected from sensors for modular actuated signal control methods. These include: (1) an immediate gap out method (instead of maintaining a gap timer); (2) a Type II dilemma zone (DZ) protection method (with the option to inhibit the extension under certain conditions if only one vehicle is present in a zone); (3) a method that identifies vehicles that are in the queue, and provides an extension until these vehicles have cleared the extension, intended to avoid split failures; (4) platoon-based secondary extension from prior research; (5) a similar control method called free optimization previously tested in the field, that uses similar concepts to provide for self-organizing coordination; and (6) an adaptive gap concepts that adjusts gap time based on existing traffic condition. Five combinations of these six methods are tested for two different detection options (full trajectory information and partial trajectory information). All methods are then compared against two conventional control methods (actuated and actuated-coordinated). The results show that the methods tend to meet their individual objectives (DZ protection, split failure mitigation, and self-organizing coordination) with varying degrees of success. These early results demonstrate that the proposed methods show promise as a component of next-generation actuated traffic signal control.