A Gear Fault Diagnosis Technique Based on ISOA-WPE-VMD and CNN-BILSTM
Abstract
In order to improve the accuracy of gearbox fault classification, a gearbox fault diagnosis model based on the ISOAWPE-VMD and CNN-BILSTM is proposed. Firstly, the northern eagle algorithm and cubic chaotic mapping as well as adaptive t-distribution variance perturbation strategy are introduced to improve the sparrow algorithm (abbreviated as ISOA). This method is introduced into VMD algorithm optimization. The weighted arrangement entropy (abbreviated as WAE) is used to screen the optimal components of VMD. For features, nine time domain features are extracted as the optimal fault features. Then CNN and BILSTM are combined to construct a CNNBILSTM fault diagnosis model. To verify the effect of the proposed ISOA-WPE-VMD and CNN-BILSTM model, a gear fault dataset of Southeast University and Huazhong University of Science and Technology are introduced for example analysis, and the results show that the algorithm proposed in this research can achieve 98.83% as well as 98.67%, which is better compared with other algorithms.