Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
ARTIFICIAL INTELLIGENCE VALVE FAILURE PREDICTION AND DIAGNOSIS SYSTEM OF AUTOMATIC VALVE
Document Type and Number:
WIPO Patent Application WO/2022/177182
Kind Code:
A1
Abstract:
An artificial intelligence valve failure diagnosis system based on edge computing for a plant automatic control valve comprises: a positioner provided at an automatic valve; a plurality of IIoT sensors provided in the automatic valve; an artificial intelligence valve failure diagnosis unit for identifying abnormal driving of the automatic valve or the type of failure by using an auto decoder deep learning algorithm to infer and analyze, in real time, a signal transmitted from the positioner and signals measured from at least some or all of the plurality of IIoT sensors; and a management center for performing data communication and remote monitoring with the failure diagnosis unit by using an Ethernet internal communication network, wherein the positioner, the IIoT sensors, and the failure diagnosis unit are provided in every individual automatic valve, and the failure diagnosis unit calculates a model value through learning and training by using data obtained during normal driving of the automatic valve, infers, by means of the model value, data obtained during abnormal driving, and determines failure if a recovery error value is larger than a threshold valve, which is the boundary value between a recovery error maximum value and a recovery error minimum valve.

Inventors:
KIM TAE JOON (KR)
Application Number:
PCT/KR2022/001098
Publication Date:
August 25, 2022
Filing Date:
January 21, 2022
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
MIRAE ENG CO LTD (KR)
International Classes:
G05B23/02; F16K37/00; G06N20/00; G16Y40/10
Foreign References:
KR20200137252A2020-12-09
KR20160019460A2016-02-19
KR102058088B12019-12-20
KR20190100830A2019-08-29
JP2017173879A2017-09-28
Attorney, Agent or Firm:
MUHANN PATENT & LAW FIRM (KR)
Download PDF: