Title:
MACHINE LEARNING DEVICE, DATA PROCESSING SYSTEM, INFERENCE DEVICE, AND MACHINE LEARNING METHOD
Document Type and Number:
WIPO Patent Application WO/2021/210497
Kind Code:
A1
Abstract:
[Problem] To accurately ascertain an abnormality and signs of an abnormality in a fluid-pressure drive valve without depending on, e.g., the experience of a worker. [Solution] A machine learning device (200) is applied to a fluid-pressure drive valve (10) that includes: a main valve (11); a drive device (12) that drives the main valve (11), and includes a cylinder and a piston; and an electromagnetic valve (1) that controls the supply and discharge of a drive fluid (A) to and from the drive device (12). The machine learning device (200) comprises: a training dataset storage unit (202) that stores a plurality of training datasets, each constituted from input data including the opening degree of the main valve (11), the pressure of the output-side drive fluid supplied to the drive device (12) from the electromagnetic valve (1), and the position of the piston (122) relative to the cylinder, and output data comprising diagnostic information for the drive valve (12) which is associated with the input data; a training unit (203) that trains a learning model for inferring the correlation between the input data and the output data by inputting a plurality of training datasets thereto; and a trained model storage unit (204) that stores the trained model.
Inventors:
AOYAMA FUMIAKI (JP)
Application Number:
PCT/JP2021/014969
Publication Date:
October 21, 2021
Filing Date:
April 09, 2021
Export Citation:
Assignee:
KANEKO SANGYO CO LTD (JP)
International Classes:
F16K31/163; F16K37/00; G05B23/02; G16Y10/25; G16Y40/10
Domestic Patent References:
| WO2019235599A1 | 2019-12-12 |
Foreign References:
| JPH0861551A | 1996-03-08 | |||
| JPH0413935A | 1992-01-17 |
Attorney, Agent or Firm:
AOYAMA, Jun et al. (JP)
Download PDF: