# US 2022/0276126 A1 · Anomaly detection system and anomaly detection method

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2022/0276126 A1
- Document ID: US20220276126A1
- Status: Published application
- Kind: A1
- Application number: 17/425,891
- Filed: 2019-02-22
- Published: 2022-09-01
- Earliest claimed: 2019-02-22
- Assignees: The Japan Steel Works, Ltd.
- Applicants: The Japan Steel Works, Ltd.
- Inventors: Tasuku OKIMOTO; Keita HIHARA; Yusuke FUKUSHI; Hideki TOMIYAMA
- Web page: https://patentgrep.com/patent/US20220276126A1

## CPC

- G01M 13/028: Acoustic or vibration analysis
- F16H 57/01: Monitoring wear or stress of gearing elements, e.g. for triggering maintenance
- F16H 2057/018: {Detection of mechanical transmission failures}
- G01M: TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- F16H: GEARING

## Abstract

An anomaly detection system (1) according to an aspect of the present invention is an anomaly detection system (1) for detecting an anomaly in a reduction drive (10), the anomaly detection system (1) including: a sensor (11_1 to 11_3) attached to a housing of the reduction drive (10), the sensor (11_1 to 11_3) being configured to detect vibration on the surface of the housing of the reduction drive (10); a statistical analysis unit (12) configured to perform a statistical analysis on the vibration detected by the sensor (11_1 to 11_3); a frequency analysis unit (14) configured to perform a frequency analysis on the vibration detected by the sensor (11_1 to 11_3); and an anomaly determination unit (17) configured to perform anomaly determination on the reduction drive (10) based on a result of the statistical analysis and a result of the frequency analysis.

## Claims

### Claim 1

An anomaly detection system for detecting an anomaly in a reduction drive, the anomaly detection system comprising:
a sensor attached to a housing of the reduction drive, the sensor being configured to detect vibration on a surface of the housing of the reduction drive;
a statistical analysis unit configured to perform a statistical analysis on the vibration detected by the sensor;
a frequency analysis unit configured to perform a frequency analysis on the vibration detected by the sensor; and
an anomaly determination unit configured to perform anomaly determination on the reduction drive based on a result of the statistical analysis and a result of the frequency analysis.

### Claim 2 (depends on 1)

The anomaly detection system according to claim 1, wherein
the statistical analysis unit performs the statistical analysis using acceleration of the vibration detected by the sensor, and
the anomaly determination unit determines presence or absence of an anomaly in the reduction drive based on a result of the statistical analysis.

### Claim 3 (depends on 1)

The anomaly detection system according to claim 1, wherein
the statistical analysis unit obtains a frequency distribution of a magnitude of acceleration of the vibration detected by the sensor, and
the anomaly determination unit determines presence or absence of an anomaly in the reduction drive based on a variance of the frequency distribution obtained.

### Claim 4 (depends on 1)

The anomaly detection system according to claim 1, wherein
the frequency analysis unit generates a power spectrum indicating a magnitude of acceleration relative to a frequency of the vibration detected by the sensor, and
the anomaly determination unit identifies a type of an anomaly in the reduction drive based on the power spectrum.

### Claim 5 (depends on 1)

The anomaly detection system according to claim 1, wherein
anomaly determination data associating a power spectrum indicating an anomaly with a type of an anomaly in the reduction drive is previously stored in the anomaly determination unit, and
the anomaly determination unit determines the type of the anomaly in the reduction drive by checking the power spectrum generated by the frequency analysis unit against the anomaly determination data previously stored.

### Claim 6 (depends on 1)

The anomaly detection system according to claim 1, wherein
the sensor comprises a plurality of sensors attached onto the surface of the housing of the reduction drive at a plurality of locations, and
the anomaly determination unit determines an anomaly location in the reduction drive based on results of the statistical analysis and results of the frequency analysis on the plurality of sensors.

### Claim 7 (depends on 1)

The anomaly detection system according to claim 1, wherein
the housing of the reduction drive has a substantially box shape,
the sensor comprises a plurality of sensors, each of the sensors being attached, near a corner of the housing, onto a corresponding one of three faces defining the corner, and
acceleration in three-axis directions near the corner of the housing is measured using each of the sensors.

### Claim 8 (depends on 7)

The anomaly detection system according to claim 7, wherein each of the sensors is attached to at least one of an upper side of a face of the reduction drive on which an input shaft is provided and an upper side of a face of the reduction drive on which an output shaft is provided.

### Claim 9

An anomaly detection method for detecting an anomaly in a reduction drive, the method comprising the steps of:
performing a statistical analysis on vibration on a surface of a housing of the reduction drive using a sensor attached to the housing of the reduction drive;
performing a frequency analysis on the vibration detected by the sensor; and
performing anomaly determination on the reduction drive based on a result of the statistical analysis and a result of the frequency analysis.
