Title:
DEVICE AND METHOD FOR MONITORING STATE OF VIBRATION OF ROTATING EQUIPMENT USING DEEP LEARNING-BASED TIME SERIES ANALYSIS
Document Type and Number:
WIPO Patent Application WO/2021/107342
Kind Code:
A1
Abstract:
Disclosed are a device and method for monitoring the state of vibration of rotating equipment using deep learning-based time series analysis. The device for monitoring the state of vibration of rotating equipment using deep learning-based time series analysis, according to one embodiment of the present application, may comprise: a data receiving unit which receives vibration data of rotating equipment from a vibration sensor installed for the rotating equipment; a data preprocessing unit which removes noise from the vibration data and extracts a characteristic variable from the noise-removed vibration data; a learning unit which generates, for the extracted characteristic variable, a predicted value associated with the vibration data, through deep learning-based time series analysis, and generates error information on the basis of the predicted value and the actual value of the vibration data; a margin setting unit which, on the basis of the error information, sets a margin for determining the presence or not of a defect associated with the vibration data; and a defect detection unit which receives new vibration data of the rotating equipment and determines the presence or not of a defect in the new vibration data on the basis of the margin.
Inventors:
KIM JI HYEOK (KR)
KIM HO SU (KR)
KIM HO SU (KR)
Application Number:
PCT/KR2020/011639
Publication Date:
June 03, 2021
Filing Date:
August 31, 2020
Export Citation:
Assignee:
WISEITECH CO LTD (KR)
International Classes:
G05B23/02; G05B19/404; G06N20/00
Foreign References:
| KR102015417B1 | 2019-08-28 | |||
| JP2013041448A | 2013-02-28 | |||
| KR101967663B1 | 2019-04-11 | |||
| KR100666452B1 | 2007-01-09 | |||
| KR20170036445A | 2017-04-03 | |||
| KR102138279B1 | 2020-07-28 |
Attorney, Agent or Firm:
YOU, Min Gyu (KR)
Download PDF: