Fault Diagnosis of Motor Rolling Bearing Based on GWO-SVM

Pengbin Li

《International Core Journal of Engineering》 5卷10期 (2019/09) Pp. 238-245

https://doi.org/10.6919/ICJE.201909_5(10).0035

摘要

The classification performance of Support Vector Machine (SVM) is greatly influenced by the characteristics of the sample and the selection of parameters of the SVM itself. Aiming at this situation, based on Shannon energy entropy, SVM and grey wolf optimization algorithm (GWO), a fault diagnosis method for motor rolling bearing based on GWO-SVM is proposed. The method adopts the fault-tolerant shannon energy entropy as the characteristic parameter, and extracts the first three IMF components as the characteristic signals by EMD decomposition, and calculates the Shannon energy entropy as the feature vector to obtain the sample set as the input of the multi-class SVM. When training SVM with samples, a new kernel function is constructed, and GWO is used to globally optimize the kernel function parameters of SVM, so that SVM can obtain the best classification performance and improve the accuracy of classification identification. Finally, the classification and identification of rolling bearing fault samples of Case Western Reserve University were carried out, and compared with other methods. The results show that the method has better reliability and classification accuracy.

關鍵字

SVM ； GWO ； SVM-GWO ； Shannon energy entropy ； Troubleshooting

參考文獻

Y.G.Lei,Z.J.He,Y.Y,Zi,et al. Fault diagnosis of rotatingmachinery based on muhiple ANFIS combination with Gas,Mechanical Systems and Signal Processing,vol.21(2007),p.2280-2294.

Keerthi S S,C.J.Lin: Asymptotic behaviors of supportvector machines with gaussian kernel,Neural Computation,vol.15(2003),p.1667-1689.

Mirjalili S, Mirjalili S M, Lewis A:Grey wolf optimizer,Advances in Engineering Software, vol.69 (2014), p.46-61．

Jayabarathi T,Raghunathan T,Adarsh B R,et al．Economic dispatch using hybrid grey wolf optimizer, Energy,vol.111(2016),p.630-641

F.Li, B.P.Tang, S.J.Dong:Maintaining embedded features based on orthogonal neighborhoods Simple fault diagnosis model, Chinese Journal of Scientific Instrume,vol.32(2011),p.621-627.