Abnormality diagnosis system and abnormality diagnosis method

Patent metadata

Field	Value
Publication number	US-2023024947-A1
Application number	US-202017758358-A
Country	US
Kind code	A1
Filing date	Dec 21, 2020
Priority date	Jan 14, 2020
Publication date	Jan 26, 2023
Grant date	—

Abstract

Provided are an abnormality diagnosis system and an abnormality diagnosis method that can prevent wrongly diagnosing equipment as having an abnormality when the equipment actually does not have an abnormality. An abnormality diagnosis system 20 comprises a sampler 21 and a calculator 24. The calculator 24 is configured to: perform first abnormality determination of whether there is an abnormality based on a result of first principal component analysis; in the case where a result of the first abnormality determination is that there is an abnormality, and perform second abnormality determination of whether there is an abnormality based on a result of second principal component analysis; and in the case where a result of the second abnormality determination is that there is an abnormality, diagnose the equipment as having an abnormality.

First claim

1 . An abnormality diagnosis system configured to perform abnormality diagnosis on equipment that performs operation repeating a fixed pattern, the abnormality diagnosis system comprising:
a sampler configured to sample time series data indicating a state of the equipment, the time series data repeating a pattern corresponding to the fixed pattern; and a calculator configured to perform principal component analysis based on the time series data sampled by the sampler, and diagnose whether the equipment has an abnormality, wherein the calculator is configured to: perform first principal component analysis on one pattern in the time series data indicating the state, and perform first abnormality determination of whether there is an abnormality based on a result of the first principal component analysis; in the case where a result of the first abnormality determination is that there is an abnormality, perform second principal component analysis collectively on a plurality of patterns successive in time including the pattern subjected to the first principal component analysis in the time series data indicating the state, and perform second abnormality determination of whether there is an abnormality based on a result of the second principal component analysis; and in the case where a result of the second abnormality determination is that there is an abnormality, diagnose the equipment as having an abnormality.

2 . The abnormality diagnosis system according to claim 1 , wherein the first abnormality determination is based on Q statistic calculated from the result of the first principal component analysis, and the second abnormality determination is based on Q statistic calculated from the result of the second principal component analysis.

3 . The abnormality diagnosis system according to claim 1 , wherein N2 <N1×M or k2<k1×M, where N1 is the number of sampling points of time series data used in the first principal component analysis by the calculator, k1 is the number of dimensions after dimensionality reduction by the first principal component analysis, N2 is the number of sampling points of time series data used in the second principal component analysis, k2 is the number of dimensions after dimensionality reduction by the second principal component analysis, and M is the number of the plurality of patterns where M≥2.

4 . The abnormality diagnosis system according to claim 1 , wherein the equipment is sizing press equipment that width-reduces a slab by driving a die using a drive motor, and
the time series data indicating the state of the equipment is a current waveform of the drive motor or an applied load waveform of the die.

5 . The abnormality diagnosis system according to claim 4 , further comprising
a data preprocessor configured to subject the time series data sampled by the sampler to preprocessing including a process of normalizing a waveform height and a process of classification according to a width reduction amount.

6 . An abnormality diagnosis method of performing abnormality diagnosis on equipment that performs operation repeating a fixed pattern, the abnormality diagnosis method comprising:
sampling time series data indicating a state of the equipment, the time series data repeating a pattern corresponding to the fixed pattern; performing first principal component analysis on one pattern in the time series data indicating the state; performing first abnormality determination of whether there is an abnormality based on a result of the first principal component analysis; performing, in the case where a result of the first abnormality determination is that there is an abnormality, second principal component analysis collectively on a plurality of patterns successive in time including the pattern subjected to the first principal component analysis in the time series data indicating the state; performing second abnormality determination of whether there is an abnormality based on a result of the second principal component analysis; and diagnosing the equipment as having an abnormality, in the case where a result of the second abnormality determination is that there is an abnormality.

7 . The abnormality diagnosis method according to claim 6 , wherein the first abnormality determination is based on Q statistic calculated from the result of the first principal component analysis, and the second abnormality determination is based on Q statistic calculated from the result of the second principal component analysis.

8 . The abnormality diagnosis method according to claim 6 , wherein N2<N1×M or k2<k1×M, where N1 is the number of sampling points of time series data used in the first principal component analysis, k1 is the number of dimensions after dimensionality reduction by the first principal component analysis, N2 is the number of sampling points of time series data used in the second principal component analysis, k2 is the number of dimensions after dimensionality reduction by the second principal component analysis, and M is the number of the plurality of patterns where M≥2.

9 . The abnormality diagnosis method according to claim 6 , wherein the equipment is sizing press equipment that width-reduces a slab by driving a die using a drive motor, and
the time series data indicating the state of the equipment is a current waveform of the drive motor or an applied load waveform of the die.

10 . The abnormality diagnosis method according to claim 9 , further comprising
subjecting the sampled time series data to preprocessing including a process of normalizing a waveform height and a process of classification according to a width reduction amount.

11 . The abnormality diagnosis system according to claim 2 , wherein N2<N1×M or k2<k1×M, where N1 is the number of sampling points of time series data used in the first principal component analysis by the calculator, k1 is the number of dimensions after dimensionality reduction by the first principal component analysis, N2 is the number of sampling points of time series data used in the second principal component analysis, k2 is the number of dimensions after dimensionality reduction by the second principal component analysis, and M is the number of the plurality of patterns where M>2.

12 . The abnormality diagnosis system according to claim 2 , wherein the equipment is sizing press equipment that width-reduces a slab by driving a die using a drive motor, and
the time series data indicating the state of the equipment is a current waveform of the drive motor or an applied load waveform of the die.

13 . The abnormality diagnosis system according to claim 3 , wherein the equipment is sizing press equipment that width-reduces a slab by driving a die using a drive motor, and
the time series data indicating the state of the equipment is a current waveform of the drive motor or an applied load waveform of the die.

14 . The abnormality diagnosis system according to claim 11 , wherein the equipment is sizing press equipment that width-reduces a slab by driving a die using a drive motor, and
the time series data indicating the state of the equipment is a current waveform of the drive motor or an applied load waveform of the die.

15 . The abnormality diagnosis method according to claim 7 , wherein N2<N1×M or k2<k1×M, where N1 is the number of sampling points of time series data used in the first principal component analysis, k1 is the number of dimensions after dimensionality reduction by the first principal component analysis, N2 is the number of sampling points of time series data used in the second principal component analysis, k2 is the number of dimen

Assignees

• Jfe Steel Corp

Inventors

• Nagano Tetsushi
• Hirata Takehide

Classifications

• G05B2223/02

  Indirect monitoring, e.g. monitoring production to detect faults of a system · CPC title

• G05B23/0221 Primary

  Preprocessing measurements, e.g. data collection rate adjustment; Standardization of measurements; Time series or signal analysis, e.g. frequency analysis or wavelets; Trustworthiness of measurements; Indexes therefor; Measurements using easily measured parameters to estimate parameters difficult to measure; Virtual sensor creation; De-noising; Sensor fusion; Unconventional preprocessing inherently present in specific fault detection methods like PCA-based methods · CPC title

• G01M99/00

  Subject matter not provided for in other groups of this subclass · CPC title

• G05B23/02 Primary

  Electric testing or monitoring · CPC title

• G05B23/024 Primary

  Quantitative history assessment, e.g. mathematical relationships between available data; Functions therefor; Principal component analysis [PCA]; Partial least square [PLS]; Statistical classifiers, e.g. Bayesian networks, linear regression or correlation analysis; Neural networks · CPC title