Device for searching and method for searching for similar breakdown cases

• US 10,204,321 B2
• Filed: 07/11/2012
• Issued: 02/12/2019
• Est. Priority Date: 07/11/2012
• Status: Active Grant

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Uchida, Takayuki, Suzuki, Hideaki, Fujiwara, Junsuke, Yuda, Shinya
• Primary Examiner(s): Lee, Cheung
• Application Number:
  US14/413,448
• Publication Number:
  US 20150161573A1
• Time in Patent Office: 2,407 Days
• CPC Class Codes: G06F 16/334 Query execution G06F16/335 takes precedence
• CPC Class Codes: G06Q 10/20 Administration of product repair or maintenance
• CPC Class Codes: G06Q 50/10 Services

Abstract

The device for searching for similar breakdown cases according to the present invention includes: a first means that detects a signal abnormality from a repair target machine, calculates the correlations between the signal abnormality and past abnormality cases by means of quantification, and regards the past abnormality cases as similar abnormality cases; a second means that obtains the component repair histories of the repair target machine in the similar abnormality cases by means of quantification; and a third means that, in accordance with the quantified correlations and the quantified component repair histories, determines and provides the priority levels of the plural past similar abnormality cases.

Claims

1. A device for searching for similar breakdown cases, comprising:

one or more sensors disposed on a repair target machine to obtain sensor data of the repair target machine; and

a computer programmed to;

detect a signal abnormality from the sensor data obtained by the one or more sensors disposed on the repair target machine;

calculate the correlations between the signal abnormality and past abnormality cases by means of quantification, and regard the past abnormality cases as similar abnormality cases;

obtain component repair histories of the repair target machine in the similar abnormality cases by means of quantification; and

in accordance with the quantified correlations and the quantified component repair histories, determine and provide priority levels of the plural past similar abnormality cases;

wherein the computer is programmed to obtain the number of replacements of components during the time period from the occurrence of a similar abnormality case in the past to the present time from an inspected component database, and calculate the ratio of the number of replacements of the components to the number of breakdown-causing component candidates stored in an abnormality-causing components database for each of the abnormality modes;

wherein the computer is programmed to weight the priority levels of similar abnormality detection cases in the past using the quantified correlations calculated by an abnormality case detection unit and the ratio between the components obtained by an inspected component search unit, and provide the weighted priority levels; and

wherein the computer is programmed to weight the quantified correlation using a larger value for an abnormality case if the abnormality case has higher correlations with past abnormality cases, and weight the ratio between the components using a larger negative value if the number of replacements of components during the time period from the occurrence of a similar abnormality case in the past to the present time is larger.

2. The device for searching for similar breakdown cases according to claim 1, further comprising:

an abnormality case database that stores abnormality detection cases that occurred in the past in association with corresponding abnormality modes;

the abnormality-causing component database that stores breakdown-causing component candidates for each of the abnormality modes; and

the inspected component database that stores the contents of the repairs/replacements of components that were performed in the past.

3. The device for searching for similar breakdown cases according to claim 2,

wherein the inspected component database stores the repair/replacement histories of the components of a machine, and the repair/replacement history of a component can be searched for from the date and time of the replacement of the component and the type of the component.

4. The device for searching for similar breakdown cases according to claim 2,

wherein the abnormality case database stores the abnormality cases of a machine that occurred in the past in association with information about the corresponding processing contents, and a processing content of the machine can be searched for from the corresponding date and time.

5. The device for searching for similar breakdown cases according to claim 2,

wherein the abnormality case database stores the abnormality cases of a machine that occurred in the past in association with information about the corresponding processing contents, and a similar abnormality can be searched for from the analysis result of the sensor data at the time of the occurrence of an abnormality.

6. The device for searching for similar breakdown cases according to claim 2,

wherein the computer is programmed to calculate the ratio of the number of repair component candidates stored in the abnormality-causing component database to the number of repaired or replaced components stored in the inspected component database.

7. The device for searching for similar breakdown cases according to claim 1,

wherein the priority levels of the plural past similar abnormality cases are arranged in descending order to be used for repairing the repair target machine.

8. A device for searching for similar breakdown cases, comprising:

one or more sensors disposed on a maintenance target machine to obtain sensor data of the maintenance target machine; and

a computer programmed to;

bring in a signal sent from the sensor data obtained by the one or more sensors disposed on the maintenance target machine, and detect an abnormality signal;

store abnormality detection cases that occurred in the past in association with corresponding abnormality modes;

store breakdown-causing component candidates for each of the abnormality modes;

store the contents of the repairs/replacements of components that were performed in the past;

calculate correlations between the abnormality signal and past abnormality cases stored in the abnormality case database by means of quantification on detecting the abnormality signal, and regard the past abnormality cases as similar abnormality cases;

obtain the number of replacements of components during the time period from the occurrence of a similar abnormality case in the past to the present time from an inspected component database, and calculate the ratio of the number of replacements of the components to the number of breakdown-causing component candidates stored in an abnormality-causing components database for each of the abnormality modes; and

weight priority levels of similar abnormality detection cases in the past using the quantified correlations calculated by an abnormality case detection unit and the ratio between the components obtained by an inspected component search unit, and provide the weighted priority levels;

wherein the computer is programmed to weight the quantified correlation using a larger value for an abnormality case if the abnormality case has higher correlations with past abnormality cases, and weight the ratio between the components using a larger negative value if the number of replacements of components during the time period from the occurrence of a similar abnormality case in the past to the present time is larger.

9. The device for searching for similar breakdown cases according to claim 8,

wherein the priority levels of the plural past similar abnormality cases are arranged in descending order to be used for repairing the repair target machine.

10. A method for searching for similar breakdown cases, the method comprising:

storing abnormality detection cases in an abnormality case database that occurred in the past in association with corresponding abnormality modes;

storing breakdown-causing component candidates in an abnormality-causing component database for each of the abnormality modes;

disposing one or more sensors on a repair target machine to obtain sensor data of the repair target machine;

detecting a signal abnormality from the sensor data obtained by the one or more sensors disposed on the repair target machine;

calculating the correlations between the signal abnormality and past abnormality cases by means of quantification, wherein the past abnormality cases are regarded as similar abnormality cases;

obtaining the number of replacements of components during the time period from the occurrence of a similar abnormality case in the past to the present time from an inspected component database, and calculating the ratio of the number of replacements of the components to the number of breakdown-causing component candidates stored in the abnormality-causing components database for each of the abnormality modes;

weighting priority levels of similar abnormality detection cases in the past using the quantified correlations calculated by an abnormality case detection unit and the ratio between the components obtained by an inspected component search unit, and providing the weighted priority levels; and

weighting the quantified correlation using a larger value for an abnormality case if the abnormality case has higher correlations with past abnormality cases, and weighting the ratio between the components using a larger negative value if the number of replacements of components during the time period from the occurrence of a similar abnormality case in the past to the present time is larger.

11. The method for searching for similar breakdown cases according to claim 10, further comprising:

using the priority levels of the plural past similar abnormality cases which are arranged in descending order for repairing the repair target machine.