Application of the Traffic Fundamental Diagram to Assess Detector Performance

This study develops new methods for evaluating detector health via event-based outputs and existing traffic flow theory. In this work, event-based detector data outputs were used to develop empirical vehicle volume-density curves per Greenshields fundamental model. Through integration, these empirical lines were compared with a conceptual volume-density curve for each detector, which was generated with average headway and posted speed limit data. The detector performance and site information were also used to model a predicted volume-density relationship for each detector on the basis of empirical observations, which was then compared with the conceptual line in the same manner as the empirical lines. The outcomes of each comparison were then used to create a database for assessing detector health within the structure of an algorithm. The algorithm is presented and discussed, followed by directions for future research, applications for practice, lessons learned, and limitations of this work.