Correcting Acceleration-distorted Inductive Signatures to Improve Traffic Information

Existing length and axle-based vehicle classification models using in-pavement sensor technologies such as inductive loops, piezo sensors and magnetometers assume constant speeds in providing estimates of vehicle length and axle gaps. However, the constant speed assumption is violated under congested stop-and-go freeway conditions and on many arterial streets where vehicles often undergo significant acceleration and deceleration. This results in erroneous length estimates which affect vehicle classification accuracy. The model developed in this study is based on a new inductive loop sensor called the Blade connected to an advanced high-speed sampling loop detector card which yields high fidelity inductive vehicle signatures. Due to its unique geometry which allows double sensors to be installed within close proximity, this technology allows the development of a new speed-profile-based temporal-spatial transformation algorithm. This new algorithm determines the change in speed as each vehicle traverses the Blade inductive sensors, subsequently correcting inductive signatures distorted by acceleration and deceleration to yield accurate vehicle length measures as well as reliable distortion-corrected vehicle signatures. The results from control vehicles indicate a significant decrease in variance of length estimates and excellent distinction of medium and heavy commercial vehicles from passenger and light commercial vehicles. The improved reliability in inductive vehicle signatures also indicate the potential of developing anonymous vehicle tracking systems through the matching of vehicle signatures at different locations even when traffic is unstable.


  • English

Media Info

  • Media Type: DVD
  • Features: Figures; References; Tables;
  • Pagination: 21p
  • Monograph Title: TRB 89th Annual Meeting Compendium of Papers DVD

Subject/Index Terms

Filing Info

  • Accession Number: 01152924
  • Record Type: Publication
  • Report/Paper Numbers: 10-0604
  • Files: BTRIS, TRIS, TRB
  • Created Date: Jan 25 2010 10:18AM