A New Deep Learning Model for Fault Diagnosis with Good Anti-Noise and Domain Adaptation Ability on Raw Vibration Signals
By Gaoliang Peng, Chuanhao Li, Yuanhang Chen, Zhujun Zhang. Italian National Conference on Sensors, 2017
1,600 citations recorded by the source corpus.
Intelligent fault diagnosis techniques have replaced time-consuming and unreliable human analysis, increasing the efficiency of fault diagnosis. Deep learning models can improve the accuracy of intelligent fault diagnosis with the help of their multilayer nonlinear mapping ability. This paper proposes a novel method named Deep Convolutional Neural Networks with Wide First-layer Kernels (WDCNN). The proposed method uses raw vibration signals as input (data augmentation is used to generate more inputs), and uses the wide kernels in the first convolutional layer for extracting features and suppressing high frequency noise. Small convolutional kernels in the preceding layers are used for multilayer nonlinear mapping. AdaBN is implemented to improve the domain adaptation ability of the model. The proposed model addresses the problem that currently, the accuracy of CNN applied to fault diagnosis is not very high. WDCNN can not only achieve 100% classification accuracy on normal signals, but also outperform the state-of-the-art DNN model which is based on frequency features under different working load and noisy environment conditions.
Subfields: Civil and Structural Engineering, Control and Systems Engineering
View Paper: https://www.semanticscholar.org/paper/e67393254aa1e30e792a3e02ef3cbdd2fb20b382
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Civil and Structural Engineering
In Control and Systems Engineering
Citations in the corpus
- Cross-Domain Fault Diagnosis of Rolling Element Bearings Using Deep Generative Neural Networks (2019, background): “[19] to improve the DA ability of neural network.”
- Deep Transfer Network with Joint Distribution Adaptation: A New Intelligent Fault Diagnosis Framework for Industry Application (2018, methodology): “The works in [20] applied the 1D CNN with wide kernels in the first convolutional layer to capture the low frequency features and restrain the high frequency noise from bearing fault signal.”
- Limited Data Rolling Bearing Fault Diagnosis With Few-Shot Learning (2019, methodology): “(1) We proposed a few-shot learning approach for bearing fault diagnosis with limited data, which is achieved by developing a siamese neural network model based on deep convolutional neural networks with wide firstlayer kernels (WDCNN) [8].”
- Potential, challenges and future directions for deep learning in prognostics and health management applications (2020, methodology): “AdaBN (Zhang et al., 2017b), adversarial training (Zhang et al., 2018; Wang et al., 2019b), and MMD-minimization (Li et al., 2018, 2019) were used to align the full source and target distributions for rotating machines.”
- Machine Learning and Deep Learning Algorithms for Bearing Fault Diagnostics - A Comprehensive Review (2019, methodology): “A novel framework WDCNN (deep CNN with wide firstlayer kernels) combined with adaptive batch normalization (AdaBN) was proposed in [159].”
- Multi-Layer domain adaptation method for rolling bearing fault diagnosis (2019, methodology): “An adaptive batch normalization method was proposed by Zhang and colleagues [41] to improve the domain adaptation ability of neural network.”
- Understanding and Learning Discriminant Features based on Multiattention 1DCNN for Wheelset Bearing Fault Diagnosis (2020, methodology): “…adaptive deep convolution neural network (ADCNN) [26] based on the 2DCNN, deep convolutional neural networks with wide first-layer kernels (WDCNN) [17], residual convolution neural network (ResCNN) [23] based on the 1DCNN, gated recurrent unit (GRU), recurrent neural network (RNN), stacked…”