TY - GEN
T1 - Bearing Fault Diagnosis Using Hyper-Laplacian Priors and Non-convex Optimization
AU - Zhao, Zhibin
AU - An, Botao
AU - Wang, Shibin
AU - Qiao, Baijie
AU - Sun, Chuang
AU - Chen, Xuefeng
N1 - Publisher Copyright:
© 2018 IEEE.
PY - 2019/1/4
Y1 - 2019/1/4
N2 - Bearing fault diagnosis is one of the most important topics in the condition-based maintenance and is also a challenging problem because of the heavy noise interference. Sparse representation methods have been proven effective to solve such a challenging problem, especially through L1 norm regularization (Laplacian). In this paper, we analyze the sparse signal model deeply in the Bayesian aspect and conclude that the distribution of the wavelet coefficients (transforming by TQWT) is well modeled by a hyper-Laplacian. However, such a prior makes the optimization problem non-convex. We adopt an effective algorithm called the generalized shrinkage algorithm (GISA) to solve this non-convex problem. Furthermore, we use the k-sparsity strategy to replace the regularization parameter for adaptivity. Finally, a simulation study and a bearing fault experiment demonstrate the performance of the proposed GISA with k-sparsity, and the comparison study shows that the hyper-Laplacian distribution can estimate the bearing fault information more accurately than the Laplacian distribution.
AB - Bearing fault diagnosis is one of the most important topics in the condition-based maintenance and is also a challenging problem because of the heavy noise interference. Sparse representation methods have been proven effective to solve such a challenging problem, especially through L1 norm regularization (Laplacian). In this paper, we analyze the sparse signal model deeply in the Bayesian aspect and conclude that the distribution of the wavelet coefficients (transforming by TQWT) is well modeled by a hyper-Laplacian. However, such a prior makes the optimization problem non-convex. We adopt an effective algorithm called the generalized shrinkage algorithm (GISA) to solve this non-convex problem. Furthermore, we use the k-sparsity strategy to replace the regularization parameter for adaptivity. Finally, a simulation study and a bearing fault experiment demonstrate the performance of the proposed GISA with k-sparsity, and the comparison study shows that the hyper-Laplacian distribution can estimate the bearing fault information more accurately than the Laplacian distribution.
KW - Bearing fault diagnosis
KW - Hyper-Laplacian prior
KW - Non-convex optimization
KW - Sparse representation
UR - https://www.scopus.com/pages/publications/85061775815
U2 - 10.1109/PHM-Chongqing.2018.00217
DO - 10.1109/PHM-Chongqing.2018.00217
M3 - 会议稿件
AN - SCOPUS:85061775815
T3 - Proceedings - 2018 Prognostics and System Health Management Conference, PHM-Chongqing 2018
SP - 1239
EP - 1244
BT - Proceedings - 2018 Prognostics and System Health Management Conference, PHM-Chongqing 2018
A2 - Ding, Ping
A2 - Li, Chuan
A2 - Yang, Shuai
A2 - Ding, Ping
A2 - Sanchez, Rene-Vinicio
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2018 Prognostics and System Health Management Conference, PHM-Chongqing 2018
Y2 - 26 October 2018 through 28 October 2018
ER -