A linear model for determining coordination of two adjacent signalized intersections

Abstract

PurposeUrban networks are usually divided into several open or closed sub‐networks. Signal coordination has been recognized as one of the most efficient methods of controlling sub‐networks that have independently optimized timing plans. However, coordinating adjacent intersections in a network is a basic prerequisite to optimizing signal‐timing plans for sub‐networks. This paper aims to develop a linear model to support decisions regarding coordination of adjacent signals.Design/methodology/approachThis paper aims to develop a linear model to support decisions regarding coordination of adjacent signals. The tests of this model which using the field data differ from those for calibration from various roadways, indicating that the model has transferability. Evaluations using microscopic simulation show that the model can objectively determine whether or not to interconnect adjacent signals depending on various traffic demands.FindingsThe model was calibrated by stepwise regression analysis with a total of 195 field samples. This model consists of the dependent variable critical block length (CL) between adjacent intersections, and the independent variables original platoon size (OPS) and platoon completeness ratio (PCR). The calibrated model is shown as following: CL = 689.97 + 6.86 OPS−7.15 PCR.Originality/valueThe proposed model appears to be a viable solution for determining whether to coordinate adjacent signals according to various traffic demands for variously configured roadways. The model shows that a larger OPS or a smaller PCR implies a larger CL. The model also indicates that adjacent signals must be interconnected if they are separated by 690 meters or less. The results also suggest that OPS from 10 to 30 fully disperse at about 800 meters downstream of a stop line. The results support the CL for effectively coordinating adjacent signals, similar to that recommended in the Manual on Uniform Traffic Control Devices. These results may be useful for the effective management of traffic signal networks.