軸承故障診斷中特徵選取的二元灰狼優化與平衡優化混合方法

A Hybrid Method of Binary Grey Wolf Optimization and Equilibrium Optimization for Feature Selection in Diagnosing Bearing Faults

李牧澤(MU-ZE LI)

指導教授 ： 許世哲 ； 李俊耀

中原大學/電機資訊學院/電機工程學系/碩士(2024年)

https://doi.org/10.6840/cycu202400225

摘要

有效擷取故障資訊並準確診斷故障仍然是軸承故障診斷的關鍵挑戰。因此，本研究提出一種診斷軸承故障的新模型。此模型主要分為三個部分：特徵擷取(feature extraction)、特徵選取(feature selection)和分類器(classifier)。在特徵擷取部分，利用經驗模態分解(empirical mode decomposition, EMD)和快速傅立葉變換(fast Fourier transform, FFT)從測試馬達的原始訊號中擷取特徵。在第二部分中，提出一種有效的特徵選取方法，並將其應用於特徵選取部分。新的特徵選取方法是基於灰狼優化(grey wolf optimizer)和平衡優化(equilibrium optimization) 最後，分類器部分利用k近鄰(k-nearest neighbor, KNN)和支持向量機(support vector machine, SVM)。為評估所提出模型的性能，應用四個軸承資料集，加州大學爾灣分校基準資料集 (University of California at Irvine, UCI)所提供之公開數據集對所提出之特徵選取方法進行驗證、馬達軸承故障電流訊號資料集、凱斯西儲大學基準資料集(Case Western Reserve University, CWRU)所和機械故障預防技術基準資料集(Machinery Failure Prevention Technology, MFPT)。實驗結果表明，與其他方法相比，該方法能夠更有效的準確診斷軸承故障，並且具有穩健性。

關鍵字

特徵選擇 ； 故障診斷模型 ； 平衡優化器 ； 灰狼優化器

並列摘要

For diagnosis of faults in bearings remains a crucial challenge in effectively extracting fault information and accurately diagnosing faults. Therefore, this study presents a new model for diagnosing faults in bearings. The model is primarily divided into three parts: feature extraction, feature selection, and classification. In the feature extraction part, features are extracted from the raw signals of the test motor using empirical mode decomposition and fast Fourier transform. In the second part, an effective feature selection method is proposed and employed in the feature selection section. The new feature selection method is based on grey wolf optimization and equilibrium optimizer. Finally, the classifier part utilizes k-nearest neighbors and support vector machine. To evaluate the performance of the proposed model, four bearing datasets are applied, namely the University of California, Irvine benchmark dataset, motor bearing fault current signal dataset, Case Western Reserve University benchmark dataset, and mechanical fault prevention Technology benchmark dataset. The experimental results demonstrate that the proposed method is more effective in accurately diagnosing bearing faults and exhibits robustness compared to other methods.

並列關鍵字

Feature selection ； Fault diagnosis model ； Equilibrium optimizer ； Grey wolf optimizer

參考文獻

[1]X. Wang, S. Lu, K. Chen, Q. Wang, and S. Zhang, “Bearing fault diagnosis of switched reluctance motor in electric vehicle powertrain via multisensor data fusion,” IEEE Transactions on Industrial Informatics, vol. 18, no. 4, pp. 2452-2464, Apr. 2022.

[2]J. Mei, Y. Zuo, C. H. Lee, and J. L. Kirtley, “Modeling and optimizing method for axial flux induction motor of electric vehicles,” IEEE Transactions on Vehicular Technology, vol. 69, no. 11, pp. 12822-12831, Nov. 2020.

[3]J. Lu, Y. Hu, G. Chen, Z. Wang, and J. Liu, “Mutual calibration of multiple current sensors with accuracy uncertainties in IPMSM drives for electric vehicles,” IEEE Transactions on Industrial Electronics, vol. 67, no. 1, pp. 69-79, Jan. 2020.

[4]Y. Guo, Y. Yang, S. Jiang, X. Jin, and Y. Wei, “Rolling bearing fault diagnosis based on successive variational mode decomposition and the EP index,” Sensors, vol. 22, no. 10, p. 3889, May 2022.

[5]Q. Xue et al., “Feature extraction using hierarchical dispersion entropy for rolling bearing fault diagnosis,” IEEE Transactions on Instrumentation and Measurement, vol. 70, pp. 1-11, Jan. 2021.