Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
GNN-BASED MASTER STATE GENERATION METHOD FOR REAL-TIME ANOMALY DETECTION
Document Type and Number:
WIPO Patent Application WO/2022/225117
Kind Code:
A1
Abstract:
The present specification discloses a method of analyzing log data output from a programmable logic controller (PLC) to generate a master state corresponding to a normal state in a repeating cycle. In addition, as data preprocessing for master state generation, a method of generating log data as graph data is disclosed. A master pattern generation method and a cycle analysis model training method according to the present specification are distinguished from the conventional art in that the methods correspond to a technology of processing machine control language (low-level language), which is difficult to be analyzed by humans, and converting same into analyzable language (high-level language), that is, a machine language processing (MLP)-based technology which can be understood by humans and analyzed by an executed machine language (language for controlling a machine) computer.

Inventors:
WANG GI NAM (KR)
PARK JUN PYO (KR)
HAN SEUNG WOO (KR)
HAN KANG HEE (KR)
JUNG MIN YOUNG (KR)
YOO SANG CHUL (KR)
YU GEUN HO (KR)
Application Number:
PCT/KR2021/016481
Publication Date:
October 27, 2022
Filing Date:
November 12, 2021
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
UDMTEK (KR)
International Classes:
G05B23/02; G05B19/05; G05B19/418; G06N3/08
Foreign References:
JP5769138B22015-08-26
CN111966076A2020-11-20
JP2001356801A2001-12-26
KR20200094514A2020-08-07
KR20200074677A2020-06-25
CN111860769A2020-10-30
Attorney, Agent or Firm:
TNI IP LAW FIRM (KR)
Download PDF: