Title:
ELEVATOR MOTOR ABNORMALITY DETECTION SYSTEM
Document Type and Number:
WIPO Patent Application WO/2022/219820
Kind Code:
A1
Abstract:
Provided is an abnormality detection system capable of constructing a learning model for determining elevator motor abnormality with higher accuracy even for an operation with little data measured under the same premise. An abnormality detection system (18) comprises a classification unit (23), a calculation unit (24), an integration unit (25), and a learning unit (28). The classification unit (23) classifies a plurality of measurement data for each operation acquired when the motor operates into a plurality of premise groups for each premise data. The calculation unit (24) calculates, for each of the premise groups, the feature quantity of the premise group on the basis of the measurement data classified into the premise group. The integration unit (25) integrates at least one of the plurality of premise groups into an integration group on the basis of the feature quantity calculated by the calculation unit (24). The learning unit (28) constructs the learning model for determining the motor abnormality on the basis of the measurement data classified into the integration group.
More Like This:
Inventors:
CHOTOKU NORIHIRO (JP)
FUJITA TAKESHI (JP)
GOTO KEI (JP)
SAKATA KOJI (JP)
KOYASU HIDENORI (JP)
SHIGA SATOSHI (JP)
FURUYA SHIRO (JP)
FUJITA TAKESHI (JP)
GOTO KEI (JP)
SAKATA KOJI (JP)
KOYASU HIDENORI (JP)
SHIGA SATOSHI (JP)
FURUYA SHIRO (JP)
Application Number:
PCT/JP2021/015764
Publication Date:
October 20, 2022
Filing Date:
April 16, 2021
Export Citation:
Assignee:
MITSUBISHI ELECTRIC BUILDING SOLUTIONS CORP (JP)
MITSUBISHI ELECTRIC CORP (JP)
MITSUBISHI ELECTRIC CORP (JP)
International Classes:
B66B3/00; B66B11/08; B66B13/14
Foreign References:
| JP2019008354A | 2019-01-17 | |||
| JP2018045302A | 2018-03-22 | |||
| JPH11203415A | 1999-07-30 | |||
| JPH11219410A | 1999-08-10 |
Attorney, Agent or Firm:
TAKADA, TAKAHASHI & PARTNERS (JP)
Download PDF: