Bearing fault diagnosis under various operation conditions using synchrosqueezing transform and improved two-dimensional convolutional neural network

Abstract

In real-world industrial applications, bearings are typically operated under variable speeds and loads depending on the production condition, which results in nonstationary vibration signals from the bearings. Synchrosqueezing transform (SST) is a method that can effectively reflect the change in frequency with time, which is suitable for processing nonstationary bearing signals. However, significant classification features are difficult to extract from time–frequency information when operation conditions such as speed and load change frequently. Hence, an improved two-dimensional convolutional neural network (CNN) named the two-dimensional multiscale cascade CNN (2D MC-CNN) is proposed for performing bearing fault diagnosis under various operating conditions. In a 2D MC-CNN, a multiscale information fusion layer is added prior to the convolutional layer of a conventional CNN to form multiscale cascade images such that sensitive bands can be acquired for fault recognition. Experiments are conducted on bearings by considering various sets of fault categories and fault severity levels under six operating conditions. The experimental results show that the proposed method effectively extracts fault-related features and demonstrates excellent diagnostic accuracy and robustness. Comparisons with the original CNN and other typically used fault diagnosis methods based on the same dataset demonstrate the effectiveness of the proposed 2D MC-CNN and bearing fault diagnosis method.

Bearings fault diagnosis method based on MAM and deep separable dilated convolutional neural network

“…In practice, the working environment of mechanical equipment is complex and changeable, and rolling bearings often work under different load conditions, so it is very important for the model to maintain high accuracy of bearing failure identification In order to verify the superiority of MAM-DSDCNN, it is compared with convolutional neural network with wide convolution kernels (WKCNN) [27], multi-scale convolutional neural network with channel attention (CA-MCNN) [28], multiscale cascade convolutional neural network (MC-CNN) [29], DSDCNN, and MAM-LeNet5. WKCNN is a CNN model with wide convolution kernel.…”

A rolling bearing fault diagnosis method based on Markov transition field and multi-scale Runge-Kutta residual network

“…Four advanced models with excellent performance, including multi-scale models and ResNet models, are selected for comparison. These models are MDCNN [34], ResNet-34 [22], MK-ResCNN [35], and MCCNN [36]. The detailed structure and implementation can be found in the references.…”

Small sample fault diagnosis for wind turbine gearbox based on lightweight multiscale convolutional neural network

“…Deng et al [23] proposed a design method for multiscale feature fusion blocks and improved the residual blocks to extract multiscale fault feature information. Zhang et al [24] proposed a two-dimensional multiscale cascaded CNN-based method to obtain sensitive wavebands for fault identification by reconstructing MC images from multiscale information. Lei et al [25] proposed a fault diagnosis method based on Markov transfer fields (MTFs) and multidimensional convolutional neural networks.…”