Multiheadway Gap-Out Logic for Actuated Control on Multilane Approaches

Abstract

The efficiency of actuated signals depends on quick detection when the queue has discharged and flow rate has dropped below the saturation rate. The traditional detection method is to measure headway between successive vehicles. In single-lane approaches, this measurement works well because the safety need for longitudinal spacing during saturation flow keeps headway variability small. However, on multilane approaches, headways are far more variable (headways near zero are common); for a low probability of premature gap-out, the critical gap has to be set extremely long so that traffic well below the saturation flow rate can hold the light green. This critical gap requirement makes multilane operations inefficient. This paper proposes a gap-out logic that is based on multiheadways, the time needed for several vehicles to pass a detector. For example, for three-lane approaches, the variability of three- or six-vehicle multiheadways is much lower than that of single headways, and this low variability enables the controller to distinguish saturation flow from lower flow more easily. Through microsimulation, intersection operations based on multiheadway logic were compared with traditional detection and lane-by-lane detection. Multiheadway logic performed best in reducing delay and cycle length; the greatest improvement occurred when traffic was heavy.