A new algorithm for detecting erroneous loop detector data
Abstract
Reduced safety and increased congestion are a major concern on U.S. roadways. Traffic Management Centers (TMC) can reduce emergency vehicle response time with prompt detection of incidents. TMCS detect incidents by applying Automatic Incident Detection (AID) algorithms to traffic data (volume, occupancy and speed). The success of AID algorithms depends heavily on the accuracy of the traffic data that is collected by Inductive Loop Detectors. This research involves the development of a new algorithm that filters out erroneous data from these loop detectors. The algorithm uses real speed, volume and occupancy data obtained from Trans-Guide, the Texas Department of Transportation's (TxDOT) ("smart highway'' project, to construct a parameter L which is characteristic of the effective lengths of vehicles on a particular lane. This parameter is then used with speed and occupancy data for that lane to estimate the volume. A large difference in the actual volume and the estimated volume is indicative of an error in measurement. Data gathered over 12 days was used to construct L for two detectors. For testing purposes, real data over 4 days was run through the algorithm. The difference in actual volume measurement and estimated volumes in most cases turned out to be less than 5 vehicles. Data showing large difference (30-40 vehicles) was examined manually and found to be erroneous. Although tests were performed on stored data, the structure of the algorithm suggests real time application. Recommendations were made on the implementation of the algorithm to Tx-DOT.
Description
Due to the character of the original source materials and the nature of batch digitization, quality control issues may be present in this document. Please report any quality issues you encounter to [email protected], referencing the URI of the item.
Includes bibliographical references (leaves 47-49).
Issued also on microfiche from Lange Micrographics.
Includes bibliographical references (leaves 47-49).
Issued also on microfiche from Lange Micrographics.
Keywords
mechanical engineering., Major mechanical engineering.