ABNORMALITY DETECTOR OF VEHICLE

To address a present situation where a technology for detecting an abnormality of a vehicle on the basis of a traveling parameter stored in a data structure of rough granularity is needed.SOLUTION: An abnormality detector that detects an abnormality of a vehicle includes a difference time-series his...

Full description

Saved in:
Bibliographic Details
Main Authors KAYAMA TAKASHI, MURAMATSU JUNYA, TOMINAGA TAKASHI, OKAJI RYOSUKE, KUCHIKI KATSUHIRO
Format Patent
LanguageEnglish
Japanese
Published 04.10.2023
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:To address a present situation where a technology for detecting an abnormality of a vehicle on the basis of a traveling parameter stored in a data structure of rough granularity is needed.SOLUTION: An abnormality detector that detects an abnormality of a vehicle includes a difference time-series histogram data calculation unit that acquires traveling condition data containing histogram data associated with a traveling parameter employed in traveling of a vehicle, and calculates difference time-series histogram data that is a difference between presently acquired histogram data and previously acquired histogram data, a time-series stress information calculation unit that calculates time-series stress information on the vehicle on the basis of the difference time-series histogram data, an abnormality detection threshold calculation unit that calculates abnormality detection thresholds on the bases of time-series stress information for multiple vehicles, and an abnormality sign prediction unit that predicts the timing when the time-series stress information on an object vehicle exceeds an abnormality determination threshold, on the basis of a transition of the time-series stress information on the object vehicle.SELECTED DRAWING: Figure 1 【課題】粒度の粗いデータ構造で保存されている走行パラメータから車両の異常を検知するための技術が必要とされている。【解決手段】車両の異常を検知する異常検知装置は、車両の走行時の走行パラメータに対応したヒストグラムデータを含む走行状態データを取得し、今回取得時のヒストグラムデータと前回取得時のヒストグラムデータの差分である差分時系列ヒストグラムデータを算出する差分時系列ヒストグラムデータ算出部と、差分時系列ヒストグラムデータに基づいて車両の時系列ストレス情報を算出する時系列ストレス情報算出部と、複数の車両の時系列ストレス情報に基づいて異常判定閾値を算出する異常判定閾値算出部と、対象車両の時系列ストレス情報の遷移に基づいて、対象車両の時系列ストレス情報が異常判定閾値を超える時期を予測する異常予兆予測部と、を備えている。【選択図】図1
Bibliography:Application Number: JP20220045256