A fan data robustness denoising method based on information theory learning

By constructing an autoencoder model based on information theory learning and utilizing a loss function that maximizes the relevant entropy and Renyi entropy, the problem of poor noise reduction effect of wind turbine data in non-Gaussian noise environment is solved, and effective noise reduction is achieved under the condition of no clean data, thereby improving the robustness and applicability of wind turbine data.

CN122286273APending Publication Date: 2026-06-26ZHEJIANG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
ZHEJIANG UNIV
Filing Date
2026-05-28
Publication Date
2026-06-26

AI Technical Summary

Technical Problem

Existing wind turbine data denoising methods are ineffective and lack robustness when clean training data is lacking and noise exhibits non-Gaussian characteristics. In particular, the noise in the wind turbine operating environment is complex, and the mean square error loss function of traditional methods, which relies on the Gaussian assumption, is not sensitive enough.

Method used

We employ an information theory-based learning approach to construct a denoising loss function that maximizes relevance entropy and a feature regularization loss function that maximizes Renyi entropy. By training an autoencoder model on noisy data, we suppress non-Gaussian noise and guide the uniform distribution of latent features, thus achieving denoising without the need for clean data.

Benefits of technology

It effectively suppresses non-Gaussian noise, improves model robustness and data detail preservation, enhances the noise reduction and generalization performance of wind turbine data, is suitable for complex noise environments, and overcomes the engineering bottleneck of difficulty in obtaining clean data.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122286273A_ABST
    Figure CN122286273A_ABST
Patent Text Reader

Abstract

This invention discloses a robust denoising method for wind turbine data based on information theory learning. The method includes: acquiring historical wind turbine data under different operating conditions; constructing an autoencoder model containing a cascaded encoder and decoder; constructing a denoising loss function and a feature regularization loss function; directly using the acquired historical wind turbine data as training data, weighting the denoising loss function and the feature regularization loss function to construct the total loss function of the autoencoder model, and iteratively training the autoencoder model using the training data itself as the reconstruction target; inputting the wind turbine data to be denoised into the trained autoencoder model for processing to obtain the denoised wind turbine data. This invention's method does not require the use of clean, noise-free data during training, and the model training process exhibits good noise robustness, enabling effective denoising of wind turbine data.
Need to check novelty before this filing date? Find Prior Art