首页|Gearbox fault identification based on lightweight multivariate multidirectional induction network
Gearbox fault identification based on lightweight multivariate multidirectional induction network
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NSTL
Elsevier
? 2022The fault diagnosis of the wind turbine gearbox is of great significance for improving the safety of the unit operation and reducing the downtime. Therefore, aiming at the contradiction between diagnostic accuracy and complexity of diagnostic model in a noisy environment, this paper studies it and proposes Lightweight multivariate and multi-directional induction network (LM-MDINet). This method designs dense separable blocks (DS- Blocks) to enhance deep feature extraction. At the same time, by decoupling the mapping relationship between the space and the channel, the amounts of parameters are reduced. In addition, a multivariate and multi-directional induction (M-MDI) layer has been added to guide the network towards the expression of effective fault information to enhance the network's ability to learn effective information. The experimental results show that the proposed method has outstanding comprehensive performance in noisy environment to compare with other methods.
Anti-noiseFault diagnosisGearboxLightweightM-MDI
Zhu X.、Wang R.、Fan Z.、Xia D.、Liu Z.、Li Z.
展开 >
Department of Power Engineering North China Electric Power University
NSTL
Elsevier
