Bearing state recognition based on sparse coding for wavelet packet frequency bands without complete prior knowledge
MA Yunfei 1,2, JIA Xisheng2, HU Qiwei2, GUO Chiming2, XING Peng2
1. Department of Armament,Noncomissioned Officer Academy of CAPF,Hangzhou 310023,China;
2. Department of Equipment Command and Management, Shijiazhuang Campus, Army Engineering University, Shijiazhuang 050003, China
Abstract:Aiming at the problem that traditional sparse coding is not accurate enough, the sparse coding on wavelet packet frequency bands was proposed. First, the wavelet packet decomposition and optimal frequency bands selection were performed on the original signal. Over-complete sparse dictionaries were trained for each optimal frequency band. The compression and reconstruction errors of each frequency band for the signal to be tested were used as the new sparse codes. To obtain final degradation features, the absolute grey relational degree of B-mode method was used here for dimension reduction. Considering that the normal running state and serious friction state of the bearing are easy to identify, an evaluation model under incomplete information based on the above two states was established, and an early warning line was set according to the threshold preset. The simulation analysis of public bearing full life data show that the new sparse coding feature has an early warning line critical point earlier than that of traditional sparse coding feature, thus the fault alarms can be issued earlier. In addition, the discrimination, noise immunity and time complexity of the new sparse coding algorithm were also studied.
马云飞1,2,贾希胜2,胡起伟2,郭驰名2,邢鹏2. 基于小波包频带稀疏编码的非完备信息条件下轴承状态识别[J]. 振动与冲击, 2021, 40(23): 288-294.
MA Yunfei 1,2, JIA Xisheng2, HU Qiwei2, GUO Chiming2, XING Peng2. Bearing state recognition based on sparse coding for wavelet packet frequency bands without complete prior knowledge. JOURNAL OF VIBRATION AND SHOCK, 2021, 40(23): 288-294.
[1] 甘茂治, 康建设, 高崎. 军用装备维修工程学[M]. 北京:国防工业出版社, 2005, 7.
[2] ZHANG S N, KANG R, HE X F and PECHT M G. China's efforts in prognostics and health management[J]. IEEE Transactions on Components and Packaging Technologies, 2008, 31(2): 509-518.
[3] 张俊宁, 张培林, 李兵, 吴定海等. 非完备先验知识下的滑动轴承摩擦状态识别[J]. 航空动力学报, 2017, 32(7): 1704-1711.
ZHANG J N, ZHANG P L, LI B, et al. Plain bearing friction state recognition without complete prior knowledge [J]. Journal of Aerospace Power, 2017, 32(7): 1704-1711.
[4] ELAD M., AHARON M. Image denoising via sparse and redundant representations over learned dictionaries [J]. IEEE Transactions on Image Processing, 2006, 15: 3736-3745.
[5] YANG S Y, WANG M, CHEN Y G, SUN Y X. Single-image super-resolution reconstruction via learned geometric dictionaries and clustered sparse coding [J]. IEEE Transactions on Image Processing, 2012, 21: 4016-4028.
[6] YANG B, LI S T. Multifocus image fusion and restoration with sparse representation [J]. IEEE Transactions on Instrumentation & Measurement, 2010, 59: 884-892.
[7] WRIGHT J, YANG A Y, GANESH A, SASTRY S S, MA Y. Robust face recognition via sparse representation [J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2009, 31: 210-227.
[8] ZHANG X P, HU N Q, HU L, CHEN L, CHENG Z. A bearing fault diagnosis method based on the low-dimensional compressed vibration signal [J]. Advances in Mechanical Engineering, 2015, 7(7): 1-12.
[9] TANG G, YANG Q, WANG H Q, LUO G G, MA J W. Sparse classification of rotating machinery faults based on compressive sensing strategy, Mechatronics, 2015, 31: 60-67.
[10] YU F J, ZHOU F. Classification of machinery vibration signals based on group sparse representation [J]. Journal of Vibroengineering, 2016, 18(3): 1540-1554.
[10] ELAD M, STARCK J L, QUERRE P, DONOHO D L. Simultaneous cartoon and texture image impainting using morphological component analysis(MCA)[J]. Applied & Computational Harmonic Analysis, 2005, 19(3): 340-358.
[11] 苗中华, 周广兴, 刘海宁, 刘成良. 基于稀疏编码的振动信号特征提取算法与实验研究[J]. 振动与冲击, 2014, 33(15): 76-81.
MIAO Z H, ZHOU G X, LIU H N, LIU C L. Study on Experiments and Feature Extraction Algorithm of vibration signals Based on Sparse Coding [J]. Journal of Vibration and Shock, 2014, 33(15): 76-81.
[12] 余建波, 刘海强, 郑小云, 周炳海等. 基于ITD与稀疏编码收缩的滚动轴承故障特征提取方法[J]. 振动与冲击, 2018, 37(19): 23-29.
YU J B, LIU H Q, ZHENG X Y, ZHOU B H, et al. Fault feature extraction method of rolling bearings based on ITD-SCS [J]. Journal of Vibration and Shock, 2018, 37(19): 23-29.
[13] 王维刚, 刘占生. 基于改进判别字典学习的故障诊断方法[J]. 振动与冲击, 2016, 35(4): 110-114.
WANG W G, LIU Z S. Fault diagnosis method based on improved discriminative dictionary learning [J]. Journal of Vibration and Shock, 2016, 35(4): 110-114.
[14] DONOHO D L. Compressed sensing[J]. IEEE Transactions on Information Theory, 2006, 52 (4 ) : 1289-1306.
[15] TROPP J A, GILBERT A C. Signal recovery from random measurements via orthogonal matching pursuit [J]. IEEE Transactions on Information Theory, 2007, 53 (12): 4655-4666.
[16] WANG J, KWON S, SHIM B. Generalized interaction matching pursuit [J]. IEEE Transactions on Signal Processing, 2012, 60(12): 6202-6216.
[17] AHARON M, ELAD M, BRUCKSTEIN A. K-SVD: An algorithm for designing overcomplete dictionaries for sparse representation [J]. IEEE Transactions on Signal Processing, 2006, 54(11): 4311-4322.
[18] MA Y F, JIA X S, HU Q W, XU D M, et al. Laplace prior-based Bayesian compressive sensing using K-SVD for vibration signal transmission and fault detection [J] Electronics, 2019, 8(5).
[19] 邵晓刚.基于矩阵灰色B型绝对关联度的人脸检测算法研
究[D]. 长春:东北师范大学, 2012.
SHAO X G. The research of the face detection algorithm taking matrix absolute grey relational degree of B-mode as core[D]. Changchun: Northeast Normal University, 2012.