Feature Extraction Method Based on Component Weighted Reconstruction and Sparse NMF for Bearing Compound Faults of In-wheel Motor

doi:10.3901/JME.2023.09.146

Abstract

Abstract: To solve the problem of poor separation and extraction for compound fault features, a fault feature extraction method based on component weighted reconstruction (CWR) and sparse non-negative matrix factorization (SNMF) is proposed for the fault of in-wheel motor bearings. Firstly, a fusion index CIH is proposed to evaluate the information of the vibration signal from multiple perspectives and adaptively select the product function (PF) component by the local mean decomposition, then perform CWR to enhance the expression of fault characteristics. Secondly, the relationship between singular values and implicit subspace of the time-frequency energy matrix of the constructed signal is analysed, and the smoothing coefficient of the constructed signal is defined based on the variance ratio of the singular values, which is used to estimate the optimal decomposition dimension of the SNMF algorithm. Finally, Itakura-Saito (IS) distance and sparse constraint are employed to establish a SNMF algorithm, which is used to decompose the time-frequency energy matrix for dimension reduction. The subspace time-domain separation signals are obtained by the inverse short-time Fourier transform, and the fault features are extracted by spectral envelope analysis. Simulation and experiment results prove that the proposed method has realized effectively the separation and extraction of compound fault features, and has certain value in engineering application.

Key words: in-wheel motor, compound fault, feature extraction, component weighted reconstruction, sparse non-negative matrix factorization

CLC Number:

TG156

XUE Hongtao, DING Dianyong, LI Ruicheng, XU Xing. Feature Extraction Method Based on Component Weighted Reconstruction and Sparse NMF for Bearing Compound Faults of In-wheel Motor[J]. Journal of Mechanical Engineering, 2023, 59(9): 146-156.

References

[1] XUE Hongtao,DING Dianyong,ZHANG Ziming,et al. A fuzzy system of operation safety assessment using multi-model linkage and multi-stage collaboration for in-wheel motor[J]. IEEE Transactions on Fuzzy Systems,2021,30(4):999-1013.
[2] HAO Yansong,SONG Liuyang,REN Bangyue,et al. Step-by-step compound faults diagnosis method for equipment based on majorization-minimization and constraint SCA[J]. IEEE-ASME Transactions on mechatronics,2019,24(6):2477-2487.
[3] MIAO Feng,ZHAO Rongzhen. A new fault diagnosis method for rotating machinery based on SCA-fastICA[J]. Mathematical Problems in Engineering,2020,2020.
[4] 王华庆,王梦阳, 宋浏阳,等. 双约束非负矩阵分解的复合故障信号分离方法[J]. 振动工程学报,2020,33(03):590-596. WANG Huaqing,WANG Mengyang,SONG Liuyang,et al. Method of compound fault signal separation using double constraints non-negative matrix factorization[J]. Journal of Vibration Engineering,2020,33(03):590-596.
[5] ISLAM M S,ZHU Yuanyuan,HOSSAIN M I,et al. Supervised single channel dual domains speech enhancement using sparse non-negative matrix factorization[J]. Digital Signal Processing,2020,100.
[6] GE Shaodi,LUO Liuhong,LI Hongjun. Orthogonal incremental non-negative matrix factorization algorithm and its application in image classification[J]. Computational & Applied Mathematics,2020,39(2).
[7] HAO Yansong,SONG Liuyang,WANG Mengyang,et al. Underdetermined source separation of bearing faults based on optimized intrinsic characteristic-scale decomposition and local non-negative matrix factorization[J]. IEEE Access,2019,7:11427-11435.
[8] LI Changning,YU Gang,FU Baochuan,et al. Fault separation and detection for compound bearing-gear fault condition based on decomposition of marginal Hilbert spectrum[J]. IEEE Access,2019,7:110518-110530.
[9] WANG Jinjiang,GAO R X,YAN Ruqiang. Integration of EEMD and ICA for wind turbine gearbox diagnosis[J]. Wind Energy,2014,17(5):757-773.
[10] YUAN Haodong,WU Nailong,CHEN Xinyuan. Mechanical compound fault analysis method based on shift invariant dictionary learning and improved fastICA algorithm[J]. Machines,2021,9(8).
[11] TANG Gang,LUO Ganggang,ZHANG Weihua,et al. Underdetermined blind source separation with variational mode decomposition for compound roller bearing fault signals[J]. Sensors,2016,16(6).
[12] 孟宗,王亚超,胡猛. 基于改进LMD和IED-SampEn的齿轮故障特征提取方法[J]. 机械工程学报,2016,52(05):169-176. MENG Zong,WANG Yachao,HU Meng. Fault feature extraction method of gear based on improved local mean decomposition and instantaneous energy distribution-sample entropy[J]. Chinese Journal of Mechanical Engineering,2016,52(05):169-176.
[13] FEI Shen,CHEN Chao,XU Jiawen,et al. A fast multi-tasking solution:NMF-theoretic co-clustering for gear fault diagnosis under variable working conditions[J]. Chinese Journal of Mechanical Engineering,2020,33(01):182-196..
[14] WANG Xu,WANG Tianyang,MING Anbo,et al. Semi-supervised hierarchical attribute representation learning via multi-layer matrix factorization for machinery fault diagnosis[J]. Mechanism and Machine Theory,2022,167.
[15] PRASETYAWAN D,TAKAMICHI N. Sensory evaluation of odor approximation using NMF with Kullback-Leibler divergence and Itakura-Saito divergence in mass spectrum space[J]. Journal of the Electrochemical Society,2020,167(16).
[16] 赵知劲,卢宏,尚俊娜. 基于约束NMF的欠定盲信号分离算法[J]. 计算机应用研究,2011,28(5):1843-1845. ZHAO Zhijin,LU Hong,SHANG Junna. Underdetermined blind signal separation algorithm based on constrained NMF[J]. Application Research of Computers,2011,28(5):1843-1845.
[17] 梁霖,栗茂林,李利邦,等. 基于非负矩阵分解的单通道故障特征分离方法[J]. 振动,测试与诊断,2016,36(05):835-840+1019. LIANG Lin,LI Maolin,LI Libang,et al. Single-channel fault feature separation method based on non-negative matrix factorization[J]. Journal of Vibration Measurement & Diagnosis,2016,36(05):835-840+1019.
[18] MENG Debiao,WANG Hongtao,YANG Shiyuan,et al. Fault analysis of wind power rolling bearing based on EMD feature extraction[J]. CMES-Computer Modeling in Engineering & Sciences,2022,130(1):543-558.