Title:
ARTIFICIAL NEURAL NETWORK TRAINING METHOD FOR PREDICTING WHETHER VEHICLE IS IN TROUBLE, METHOD FOR DETERMINING WHETHER VEHICLE IS IN TROUBLE, AND COMPUTING SYSTEM PERFORMING SAME
Document Type and Number:
WIPO Patent Application WO/2022/114453
Kind Code:
A1
Abstract:
Disposed are a method and system for training an artificial neural network by using driving record data and trouble diagnosis data, which may be collected while a vehicle is running, and predicting, in advance, trouble of the vehicle by using same. According to one aspect of the present invention, provided is an artificial neural network training method comprising the steps of: collecting, by a computing system, driving record time series data indicating a vehicle state or driving state of a collection target vehicle during a predetermined measurement period; collecting, by the computing system, a diagnostic trouble code (DTC) generated while the collection target vehicle is running during the measurement period; generating, by the computing system, a training data set, on the basis of the driving record time series data and the diagnostic trouble code of the collection target vehicle, which are collected during the measurement period; and training, by the computing system, an artificial neural network on the basis of the generated training data set.
Inventors:
CHOI GEO SEOK (KR)
CHOI MOON KYU (KR)
PYO GEON HO (KR)
LEE GYU HYEON (KR)
CHOI MOON KYU (KR)
PYO GEON HO (KR)
LEE GYU HYEON (KR)
Application Number:
PCT/KR2021/010722
Publication Date:
June 02, 2022
Filing Date:
August 12, 2021
Export Citation:
Assignee:
INFOCAR CO LTD (KR)
International Classes:
G06N3/08; G06N3/04; G07C5/08; G07C5/10
Foreign References:
| KR101889049B1 | 2018-08-21 | |||
| KR20120107774A | 2012-10-04 | |||
| KR20180105850A | 2018-10-01 | |||
| KR20190104267A | 2019-09-09 | |||
| JP2020096286A | 2020-06-18 | |||
| KR20170076131A | 2017-07-04 |
Other References:
See also references of EP 4033413A4
Attorney, Agent or Firm:
SHIM, Choong Sup (KR)
Download PDF: