Research on Fault Diagnosis of Rolling Bearing Based on SOM Neural Network

Xiaoqiang Yao ； Gang Tang ； Xiong Hu

《International Core Journal of Engineering》 5卷9期 (2019/08) Pp. 159-166

https://doi.org/10.6919/ICJE.201908_5(9).0024

摘要

In order to solve the problems of rolling bearing fault diagnosis and complex nonlinear pattern classification, according to the non-stationary and nonlinear characteristics of rolling bearing vibration signals, by analyzing the influence of bearing vibration signal training sample set, feature extraction methods and other factors on bearing fault diagnosis, a state clustering analysis method based on Self-Organizing Maps neural network is designed. In this method, five fault features of rolling bearings are extracted, and SOM neural network is used for training to obtain a bearing state discrimination model. Finally, the trained model is used to identify the fault types of rolling bearings. The verification of actual bearing fault data shows that the fault diagnosis method based on SOM neural network can accurately and effectively identify the fault of rolling bearing, and the accuracy rate is as high as 100%.

關鍵字

Rolling bearing ； fault diagnosis ； SOM neural network ； feature extraction

參考文獻

Nandi S, Toliyat H A, Li X. Condition monitoring and fault diagnosis of electrical motors-a review[J]. IEEE Transactions on Energy Conversion, 2005, 20(4):719-729.

Li B, Zhang Y. Supervised locally linear embedding projection (SLLEP) for machinery fault diagnosis[J]. Mechanical Systems & Signal Processing, 2011, 25(8):3125-3134.

Lei Y, He Z, Zi Y. A new approach to intelligent fault diagnosis of rotating machinery[J]. Expert Systems with Applications, 2008, 35(4):1593-1600.

Mechefske C K, Mathew J. Fault detection and diagnosis in low speed rolling element bearings Part I: The use of parametric spectra[J]. Mechanical Systems & Signal Processing, 1992, 6(4):297-307.

Germen E, Başaran M, Fidan M. Sound based induction motor fault diagnosis using Kohonen self-organizing map[J]. Mechanical Systems & Signal Processing, 2014, 46(1):45-58.