Automated Identification of Traffic Detector Malfunctions
-
2022-09-01
-
Details:
-
Creators:
-
Corporate Creators:
-
Corporate Contributors:
-
Subject/TRT Terms:
-
Publication/ Report Number:
-
Resource Type:
-
Geographical Coverage:
-
Edition:Final Report
-
Corporate Publisher:
-
Abstract:There is a need to improve signalized intersection operations through identifying malfunctioning detectors, as recent work has shown that errors in data quality and accuracy may be widespread due to issues with aging equipment and unmet maintenance needs. Accordingly, there is a desire for policies, procedures, and techniques to identify malfunctioning detection equipment and evaluate the quality of data produced by detectors. Current tools, including those available through newer Advanced Traffic Controller (ATC) standards, are able to detect major detector failures by examining the presence, absence, or frequency of data being sent by a detector, but these tools are not able to assess the quality of the information sent; therefore, the health of the detector is commonly unmonitored. To address this issue, using event-based data, researchers isolated saturated flow from individual presence detectors at signalized intersections, and using detector activations and occupancy, were able to calculate volume and density metrics for individual green intervals. These data points were then used to approximate the undersaturated portion of a Volume vs. Density fundamental diagram for detectors at various sites in Oregon. Using the mathematical concept of percent difference between integrals, several performance datasets were developed for this work for use in algorithms also developed as part of this work. These algorithms assess detector health, both initially and over time. Finally, a system design and implementation plan was developed to aid in the deployment of this system.
-
Format:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type: