Thermal power generating unit fault monitoring method based on domain adversarial auto-encoder model

By constructing a domain-adversarial autoencoder model, the problem of data imbalance under small sample modes in thermal power unit fault monitoring was solved, achieving high-precision fault detection and low false alarm rate, and improving the model's adaptability and generalization ability.

CN120871805APending Publication Date: 2025-10-31SHANGHAI MINGHUA ELECTRIC POWER TECH & ENG +1
View PDF 8 Cites 0 Cited by

Patent Information

Application Number
CN202510823033.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-06-19
Publication Date
2025-10-31

Smart Images

  • Figure CN120871805A_ABST
    Figure CN120871805A_ABST
Patent Text Reader

Abstract

The invention relates to a thermal power generating unit fault monitoring method based on a domain adversarial auto-encoder model, and the method comprises the following steps: S1, collecting and storing the normal data of a historical mode and a small sample mode in the operation process of a thermal power generating unit, and carrying out the preprocessing of the data; s2, constructing a domain adversarial auto-encoder model and training the model by using the preprocessed historical data to obtain an SPE statistical magnitude control limit of the model under a preset confidence interval; S3, obtaining real-time operation data of a thermal power generating unit on line and preprocessing the operation data; and S4, taking the preprocessed real-time data as input data of the domain adversarial auto-encoder model, obtaining SPE statistics in real time, if the SPE statistics are greater than the SPE statistics, taking the data as fault data, and otherwise, taking the data as normal data. Compared with the prior art, the method has the advantages of high fault monitoring precision, high adaptability and the like.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Intelligent fault diagnosis method based on unsupervised domain adaptation

    CN114398992A

  • Multi-modal industrial process fault diagnosis method based on confrontation of local maximum mean value difference

    CN114879628A

  • Wind turbine generator abnormal state identification method and system based on improved confrontation automatic encoder

    CN117195121A

  • Equipment fault diagnosis method based on adversarial transfer learning and class balance loss

    CN117312950A

  • Unsupervised generator bearing fault detection method based on variational auto-encoder

    CN118310746A