Implementation of Self-Organizing Logic in a Field-Deployable Controller: Software-in-the-Loop Evaluation

Abstract

Traffic signal control algorithms were developed for self-organizing control using secondary extensions that can be implemented using logic features available in modern commercial traffic signal controllers. Self-organizing control is an alternative to coordinated control, using secondary extension logic to provide dynamic coordination for arriving platoons within a fully actuated, cycle-free framework. The secondary extension logic uses detectors located up to 900 ft upstream of the stop bar to detect arriving platoons and extend the green until the platoon passes, if certain platoon characteristics and intersection operation conditions are met, creating a mechanism for traffic signals to coordinate more naturally. The goal of this approach is to estimate the benefits of self-organizing control without requiring the trajectory data and custom controller firmware proposed in earlier studies. The developed algorithms were tested in a microsimulation environment using software-in-the-loop on two real corridors in Virginia, US. The first corridor is in a relatively urban area with coordinated-actuated signal control, while the second corridor is in a suburban area with largely spaced intersections that are operating under fully actuated control. Compared with coordinated-actuated control, fully actuated control was found to reduce intersection delay by approximately 25%, and adding secondary extension logic further reduced intersection delay by approximately 5%. Compared with fully actuated control, adding secondary extension logic reduced mainline delay by as much as 30%, showing the effectiveness of secondary extension logic in promoting coordination. Sensitivity analysis revealed that the developed logic likewise outperformed the other signal control alternatives under different demand scenarios.