Prediction method, device, readable medium and electronic device

The method addresses data security and accuracy issues in equipment failure prediction by using federated learning to transfer unshared data, ensuring secure and precise equipment failure and status prediction.

JP7783832B2Active Publication Date: 2025-12-10ENNEW DIGITAL TECH CO LTD
View PDF 7 Cites 0 Cited by

Patent Information

Application Number
JP2022564514
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Priority Date
2020-10-29
Filing Date
2021-06-21
Publication Date
2025-12-10
Estimated Expiration
2041-06-21

AI Technical Summary

Technical Problem

Existing equipment failure prediction methods in industrial settings face challenges due to data security issues when sharing historical operation data and tag data between devices, and variations in data distribution lead to low accuracy in equipment status prediction models.

Method used

A method and apparatus that utilize federated learning to establish a relationship between target equipment and unshared data, determining probability distribution models and weights to transfer non-shared data for equipment failure prediction, ensuring data security by avoiding direct data sharing.

Benefits of technology

Ensures data security while accurately predicting equipment failure and status by establishing a federated learning model based on unshared data, enhancing prediction accuracy and maintaining data privacy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007783832000004
    Figure 0007783832000004
  • Figure 0007783832000005
    Figure 0007783832000005
  • Figure 0007783832000006
    Figure 0007783832000006
Patent Text Reader

Abstract

Provided are a prediction method, device, readable medium, and electronic device that can transfer non-shared data to a target device, establish a relationship between feature data of the target device and device failure, eliminate the need to share feature data between devices, and ensure data security. [Solution] The present invention discloses a prediction method, apparatus, readable medium, and electronic device, which includes the steps of determining feature information of a target device and detection point data information corresponding to the target device, determining a probability distribution model of the feature data of the target device and a probability distribution model of the detection point data having non-shared data based on the feature information of the target device and the detection point data information corresponding to the target device, determining a weight of the non-shared data based on the probability distribution model of the feature data and the probability distribution model of the detection point data, establishing a federated learning model based on the non-shared data, the weight of the non-shared data, and an equipment failure tag corresponding to the non-shared data, and performing equipment failure prediction of the target device based on the federated learning model.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Rolling bearing fault diagnosis method based on multi-source weighted ensemble transfer learning

    CN111506862A

  • Mechanical intelligent fault diagnosis method based on federated learning

    CN111678696A

  • Robot and robot management device

    JP2005111603A

  • Learning device, learning method and program

    JP2010092266A

  • Improved font recognition using triplet loss neural network training

    JP2019083002A