公开(公告)号：US20190193741A1

公开(公告)日：2019-06-27

申请号：US16227465

申请日：2018-12-20

Applicant: DENSO CORPORATION ,  NATIONAL UNIVERSITY CORPORATION NARA INSTITUTE OF SCIENCE AND TECHNOLOGY

Inventor： Kentaro HITOMI ,  Kazuhito TAKENAKA ,  Masataka MORI ,  Kazushi IKEDA ,  Takatomi KUBO ,  Hiroaki SASAKI

IPC: B60W40/06 ,  G06K9/00 ,  H04W4/46 ,  G08G1/0967

CPC classification number: B60W40/06 ,  G06K9/00791 ,  G08G1/096791 ,  H04W4/46

Abstract: In an apparatus for detecting an anomaly of an evaluation target condition of a road, a storage stores a reference data set correlating to a predetermined attribute and being comprised of reference data samples. An anomaly level calculator obtains, from target vehicles, an evaluation data set correlating to the predetermined attribute and being comprised of evaluation data samples. The evaluation data samples are collected from the respective target vehicles as target driving data items at the predetermined attribute under the evaluation target condition of the road. The target driving data item for each of the target vehicles represents at least one driving operation of the corresponding one of the target vehicles. The anomaly level calculator compares the reference data set with the evaluation data set to thereby calculate an anomaly level of the evaluation target condition of the road.

公开(公告)号：US20190197038A1

公开(公告)日：2019-06-27

申请号：US16227563

申请日：2018-12-20

Applicant: DENSO CORPORATION

Inventor： Masataka MORI ,  Kentaro HITOMI ,  Kazuhito TAKENAKA

IPC: G06F16/2458 ,  G06N7/00 ,  G06F16/28 ,  G06K9/62

CPC classification number: G06F16/2462 ,  G06F16/28 ,  G06K9/6215 ,  G06N7/005

Abstract: A computer calculates, in accordance with a maximum mean discrepancy, a similarity level between a first feature distribution correlating to a first distribution information item stored in a distribution database and a second feature distribution correlating to a second distribution information item stored in the distribution database. The second distribution information item is different from the first distribution information item. The maximum mean discrepancy is a distance measure indicative of the similarity level between the first and second feature distributions. The computer determines whether the calculated similarity level is equal to or higher than a predetermined threshold, and integrates the first feature distribution and the second feature distribution into a common feature distribution upon determining that the calculated similarity level is equal to or higher than the predetermined threshold.

公开(公告)号：US20180018871A1

公开(公告)日：2018-01-18

申请号：US15646867

申请日：2017-07-11

Applicant: DENSO CORPORATION

Inventor： Kazuhito TAKENAKA ,  Masumi EGAWA ,  Utsushi SAKAI ,  Kentaro HITOMI ,  Yuki SHINOHARA ,  Hideaki MISAWA ,  Masataka MORI

IPC: G08G1/0962 ,  G07C5/02 ,  G08G1/01

CPC classification number: G08G1/0962 ,  G07C5/02 ,  G08G1/0112 ,  G08G1/0129 ,  G08G1/096716 ,  G08G1/096741 ,  G08G1/096775

Abstract: In a driving assist system, a driving evaluator compares a driving feature data item sampled at each predetermined sampling point and obtained from a target vehicle with historical driving data items for the corresponding sampling point. The driving evaluator obtains, based on a result of the comparison, an evaluation value of the driving feature data item for the target vehicle at each predetermined sampling point. An unusual driving determiner obtains a cumulative sum of selected values in the evaluation values of the driving feature data items for the target vehicle. The unusual driving determiner determines whether the cumulative sum is larger than a predetermined threshold, and determines that driving of a driver of the target vehicle is unusual upon determining that the cumulative sum is larger than the predetermined threshold.

公开(公告)号：US20180023965A1

公开(公告)日：2018-01-25

申请号：US15658199

申请日：2017-07-24

Applicant: DENSO CORPORATION

Inventor： Hideaki MISAWA ,  Masumi EGAWA ,  Utsushi SAKAI ,  Kentaro HITOMI ,  Yuki SHINOHARA ,  Kazuhito TAKENAKA ,  Masataka MORI

IPC: G01C21/34 ,  G01C21/36 ,  G06N5/04

Abstract: An anomaly estimation apparatus includes a collection section that collects vehicle data, a feature amount calculation section that calculates a feature amount from the vehicle data and stores the feature amount and a place corresponding thereto, an anomaly determination section that determines whether an anomaly occurrence point is present based on the feature amount, an accumulation section that, if the anomaly occurrence point is present, uses the vehicle data at the anomaly occurrence point and an anomaly periphery point to generate estimation data, an information generation section that uses the estimation data to generate causality information representing causality between an anomaly caused at the anomaly occurrence point and an anomaly caused at the anomaly periphery point, and an estimation section that, if the anomaly occurrence point is present, uses the causality information to estimate transition of the anomaly from the anomaly occurrence point to the anomaly periphery point.

公开(公告)号：US20170337812A1

公开(公告)日：2017-11-23

申请号：US15589812

申请日：2017-05-08

Applicant: DENSO CORPORATION

Inventor： Masataka MORI ,  Hideaki MISAWA ,  Kazuhito TAKENAKA ,  Yuki SHINOHARA ,  Kentaro HITOMI ,  Utsushi SAKAI ,  Masumi EGAWA ,  Kenji MUTO

IPC: G08G1/01 ,  G06F17/30

CPC classification number: G08G1/0133 ,  G06F16/29 ,  G08G1/0112 ,  G08G1/0141

Abstract: In a driving-state data storage apparatus, a collector collects, from each of vehicles on a target travelling road, a value of data indicative of a driving state of the corresponding vehicle to correspondingly obtain driving-state data values for the target road. A data allocator divides, based on similarity among the driving-state data values, the target traveling road into a plurality of traveling segments, and extracts, from the driving-state data values, data values for each of the divided travelling segments. The data values extracted for each of the travelling segments are similar to each other. The data allocator allocates a distribution of the extracted data values for each of the divided travelling segments to the corresponding one of the divided travelling segments as a feature distribution. A storage unit stores the feature distribution allocated for each of the travelling segments.