SAFETYLIT WEEKLY UPDATE

We compile citations and summaries of about 400 new articles every week.
RSS Feed

HELP: Tutorials | FAQ
CONTACT US: Contact info

Search Results

Journal Article

Citation

Ikefuji D, Prasad H, Matsushita T, Sakurai H, Suzuki T, Matsuda Y, Otani M. Proc. Conf. Japan Soc. Traffic Eng. 2022; 42: 487-490.

Vernacular Title

光ファイバセンシング技術を用いた高速道路の低速事象検知

Copyright

(Copyright © 2022, Japan Society of Traffic Engineers)

DOI

10.14954/jsteproceeding.42.0_487

PMID

unavailable

Abstract

We have so far proposed a method to capture the traffic flow from a bird's-eye view with high accuracy by extracting the trajectories of traveling vehicles on expressways using optical fiber sensing technology and analyzing them in time spans of the order of minutes. However, if the processing time width is shortened in order to immediately detect abnormalities on expressways, such as low-speed events, the amount of noise in the received signal increases, resulting in false speed estimates and frequent false alarms. There is Therefore, in this paper, we propose a method to detect low-speed events in the order of tens of seconds by improving speed estimation accuracy by weighting extracted trajectories and by defining a low-speed event model to suppress false alarms. As a result of evaluation using actual data including low-speed events due to traffic jams, etc., we achieved detection accuracy of 90% or more for low-speed events and a false alarm rate of 10% or less with a processing time of 30 seconds.

===

我々はこれまでに、光ファイバセンシング技術を利用して高速道路上の走行車両軌跡を抽出し、分オーダーの時間幅で分析することで交通流を俯瞰で高精度に捉える手法を提案してきた。しかし、低速事象のような高速道路上の異常を即時検知するために処理時間幅を短くした場合、受信信号のノイズ量が相対的に増加することで速度誤推定が生じ、誤報が多発する課題がある。そこで本稿では、抽出軌跡の重み付けによる速度推定精度の向上、および低速事象のモデルを定義することで誤報を抑圧し、数十秒オーダーで低速事象を検知する手法を提案する。渋滞等による低速事象を含む実データで評価を実施した結果、30 秒の処理時間で低速事象の検出精度 90%以上、誤報率 10%以下を達成した。


Language: ja

Keywords

交通流; 低速検知; 光ファイバセンシング; 全線俯瞰; 即時検知

NEW SEARCH


All SafetyLit records are available for automatic download to Zotero & Mendeley
Print