Skip to Main content Skip to Navigation
Journal articles

A Wavelet Threshold Denoising-based Imbalance Fault Detection Method for Marine Current Turbines

Abstract : Blade imbalance fault caused by the marine organisms is considered as the most importantfault in marine current turbines. Therefore, it is important to accurately detect the fault in time to correct it,minimize the downtime, and maximize the productivity. Imbalance fault detection methods using generatorstator current signals have attracted attentions due to their low cost, operability and stability compared tothe ones using vibration analysis. However, it is difficult to extract the fault signature and automaticallydetect the imbalance fault under different flow velocity conditions. In this paper, a wavelet thresholddenoising-based imbalance fault detection method using the stator current is proposed. The signal isanalyzed through three consecutive steps: the parameters offline setting based on wavelet thresholddenoising, the Hilbert transform method and the Principle Component Analysis-based detection method.With this method, the imbalance fault can be detected automatically. The proposed approach of imbalancefault detection is assessed under different flow velocity conditions and validated using an experimentalplatform. The results are promising with false alarm and false negatives rates less than 1% and 5%respectively when using Q statistic. Moreover, the experimental results show that the proposed method hasgood stability under different flow velocity conditions.
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-02479000
Contributor : Sandrine Charlier <>
Submitted on : Friday, February 14, 2020 - 11:40:41 AM
Last modification on : Wednesday, April 1, 2020 - 1:59:01 AM

Links full text

Identifiers

Citation

Zhichao Li, Tianzhen Wang, Yide Wang, Yassine Amirat, Mohamed Benbouzid, et al.. A Wavelet Threshold Denoising-based Imbalance Fault Detection Method for Marine Current Turbines. IEEE Access, IEEE, 2020, 8, pp.29815-29825. ⟨10.1109/ACCESS.2020.2972935⟩. ⟨hal-02479000⟩

Share

Metrics

Record views

67