Author: Ho Lee

Publisher:

Published: 2007

Total Pages: 214

ISBN-13:

Abstract: A Freeway Management System (FMS) acquires data from the roadway and processes these data to identify and respond to problems, notifying operators and motorists of those problems. If some aspects of the data collection are unreliable, then the response decisions and the information given may well be faulty. Hence, accurate traffic data acquisition is essential for effective traffic surveillance and subsequent management applications. Loop detectors, the most commonly used vehicle detectors for freeway traffic surveillance, are not always calibrated correctly, so it is necessary to identify potentially inaccurate detectors. This thesis presents an evaluation of the performance of the loop detectors on 1-71 in Columbus, Ohio. The evaluation includes the percentage of vehicles actuating only one loop in a dual loop detector, and detector mapping error tracked by the relationship of speed and occupancy from a dual loop detector. In addition, loop correction factors for both single and dual loop detectors are calculated to improve the accuracy of speed estimates and measurements. The analysis employs both statistical trends gathered from the detectors and concurrent velocities collected from probe vehicles as they pass over the detectors. As shown herein, loop detector's sensitivity can change over time which impacts speed and occupancy from that loop. So, the trend of daily median speed for off-peak time periods is used to determine the change in sensitivity of loop detectors over long time periods. This trend is then used to illustrate the fact that the correction factors can abruptly change. In the course of this work performance measurements of the freeway using the corrected loop detector data are developed, namely average daily traffic and delay. The weekday median for both of the daily measures is calculated at each week to track trends over years. The weekly average daily traffic and delay present an overview of the freeway system usage and performance. Also, summary plots and summary difference plots are developed to show how traffic condition evolves over time and space for identifying traffic condition and recurring congestion. Although presented in the context of the Columbus system, the tools should be generalizable to most freeway surveillance systems.