ABNORMALITY DETECTION METHOD, ABNORMALITY DETECTION APPARATUS, AND ABNORMALITY DETECTION SYSTEM

To provide an abnormality detection method, an abnormality detection apparatus and an abnormality detection system capable of bringing a detection accuracy of presence or absence of abnormality related to behavior of a vehicle closer to a predetermined detection accuracy.SOLUTION: When detecting whe...

Full description

Saved in:
Bibliographic Details
Main Authors MORI MAKI, MUTO KENJI
Format Patent
LanguageEnglish
Japanese
Published 03.12.2020
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:To provide an abnormality detection method, an abnormality detection apparatus and an abnormality detection system capable of bringing a detection accuracy of presence or absence of abnormality related to behavior of a vehicle closer to a predetermined detection accuracy.SOLUTION: When detecting whether abnormality is found in behavior of a vehicle depending on a magnitude of a degree of abnormality that indicates a degree of difference between first driving information representing behavior of a vehicle which is a target for detecting abnormality and second driving information in a case where abnormality is not found in a traveling environment using a computer, a degree of abnormality is estimated as an abnormality threshold so that a probability of erroneous determination related to presence or absence of abnormality by using a degree of abnormality distribution 4 generated from a group of degrees of abnormality stored in advance, and it is detected whether abnormality is found in the behavior of the vehicle by comparing the estimated abnormality threshold and the degree of abnormality for the first driving information.SELECTED DRAWING: Figure 5 【課題】車両の挙動に関する異常の有無の検出精度を所定の検出精度に近づけることができる異常検出方法、異常検出装置、及び異常検出システムを提供する。【解決手段】コンピュータを用いて、異常の検出対象となる車両の挙動を表す第1の運転情報と、走行環境に異常が認められない場合における第2の運転情報の相違度合いを示す異常度の大きさに応じて、車両の挙動に異常が認められるか検出する場合に、予め蓄積された異常度の集合から生成される異常度分布4を用いて、異常の有無に関する誤判定の確率が予め定めた確率となるような異常度を異常閾値として推定し、推定した異常閾値と、第1の運転情報に対する異常度を比較することで、車両の挙動に異常が認められるか検出する。【選択図】図5
Bibliography:Application Number: JP20190101236