Abstract:Accurate fault diagnosis is vital for modern maintenance strategies to improve machinery reliability and efficiency. Automated predictive tools, such as deep learning, are gaining more attention as the need for more general and robust diagnosis algorithms is crucial. In this work, a rotational-speed-independent diagnosis algorithm based on using a novel 2D color-coded map as the input to a deep artificial neural network (DNN) is proposed. The 2D map is named normalized diagnostic feature-gram (NDFgram). The proposed algorithm is applied for bearing fault diagnosis to investigate its effectiveness. For that purpose, the bearing vibration signals are processed first to obtain the bi-frequency spectral coherence data. Secondly, diagnostic features (DF) are calculated at specific cyclic frequencies owing to bearing faults by integrating the obtained SCoh data over the spectral frequency domain using a center frequency and frequency range. The calculated DFs are represented by a 2D map against the center frequency and frequency resolution. The maps from different fault features are stacked together to form the diagnostic patterns. Thirdly, a pretrained convolutional neural network (CNN) is applied to learn the feature pattern and diagnose the bearing faults. The CNN is trained using fixed-speed data and then it is applied to diagnose faults in the test data recorded at the same speed. Then, it is also tested using variable-speed data and data of another ball bearing type in order to show the independency on the rotational speed and ball bearing type in practice. The results show a 100% success rate for the constant-speed tests and 98.16% accuracy for the variable-speed testing dataset. The accuracy of diagnosing the faults of the second type of ball bearing is 98.56%. The diagnosis accuracy of the proposed method is still high even when a white noise is artificially added to the signals in the noise insusceptibility test.

Bearing Fault Diagnosis Using Lightweight and Robust One-Dimensional Convolution Neural Network in the Frequency Domain

Time-Frequency Multi-Domain 1D Convolutional Neural Network with Channel-Spatial Attention for Noise-Robust Bearing Fault Diagnosis

Intelligent Bearing Fault Diagnosis Based on Feature Fusion of One-Dimensional Dilated CNN and Multi-Domain Signal Processing

Bearing Fault Diagnosis using Normalized Diagnostic Feature-gram and Convolutional Neural Network

Rolling Bearing Fault Diagnosis based on Correlated Channel Attention-Optimized Convolutional Neural Networks

A multi-scale convolutional neural network for bearing compound fault diagnosis under various noise conditions

A multiscale convolution neural network for bearing fault diagnosis based on frequency division denoising under complex noise conditions

Rolling Bearing Fault Diagnosis Based on Convolutional Neural Network and Multi-sensor Information Fusion

A Novel Anti-Noise Fault Diagnosis Approach for Rolling Bearings Based on Convolutional Neural Network Fusing Frequency Domain Feature Matching Algorithm

Fault Diagnosis of Rotating Machinery under Noisy Environment Conditions Based on a 1-D Convolutional Autoencoder and 1-D Convolutional Neural Network

An Intelligent Fault Diagnosis Method of Rolling Bearings Based on Short-Time Fourier Transform and Convolutional Neural Network

A Bearing Fault Diagnosis Method Based on Spectrum Map Information Fusion and Convolutional Neural Network

A Generic Intelligent Bearing Fault Diagnosis System Using Compact Adaptive 1D CNN Classifier

Bearing Fault Diagnosis with DDCNN Based on Intelligent Feature Fusion Strategy in Strong Noise

Intelligent Fault Diagnosis of Rolling Element Bearings Based on Modified AlexNet

A new bearing fault diagnosis method based on modified convolutional neural networks

An Anti-Noise Convolutional Neural Network for Bearing Fault Diagnosis Based on Multi-Channel Data

Deep Residual Multiscale Convolutional Neural Network With Attention Mechanism for Bearing Fault Diagnosis Under Strong Noise Environment

Rolling bearing fault diagnosis based on Efficient Time Channel Attention optimized deep Multi-scale Convolutional Neural Networks

Research on Intelligent Fault Diagnosis Method for Rolling Bearing Based on One-Dimensional LeNet-5 Convolutional Neural Network

Intelligent Fault Diagnosis of Bearing Based on Convolutional Neural Network and Bidirectional Long Short-Term Memory