Bearing Fault Research Articles

Subway Train Bearing Fault Detection Under Variable Speed Based on Speed-Guided ACMD and Order Tracking

Subway Train Bearing Fault Detection Under Variable Speed Based on Speed-Guided ACMD and Order Tracking

Unsupervised Contrastive Learning-Based Single Domain Generalization Method for Intelligent Bearing Fault Diagnosis

Unsupervised Contrastive Learning-Based Single Domain Generalization Method for Intelligent Bearing Fault Diagnosis

Period oriented weighting sparse denoising for compound faults of rolling bearing

Period oriented weighting sparse denoising for compound faults of rolling bearing

Integration of Bayesian optimization into hyperparameter tuning of the particle swarm optimization algorithm to enhance neural networks in bearing failure classification

Bearings are a primary source of defects in induction motors (IM), requiring effective diagnostic methods. Neural Networks (NNs) are useful for this purpose, but optimizing their parameters is challenging. This study presents an alternative approach that modifies Particle Swarm Optimization (PSO) by integrating Bayesian optimization to dynamically tune PSO hyperparameters, enhancing the NN’s ability to detect defects in IM bearings. Unlike traditional methods with empirically determined hyperparameters, this approach adapts to varying data conditions for better performance. The method was tested using vibration and current signals of different durations (2s, 1s, 0.5s, 0.25s) and torque ranges (0Nm to 22Nm) from a laboratory-generated dataset, with results compared to those obtained using other optimizers. The accuracy achieved was 92.57% for vibration signals at 0.25s and 97.23% across torque ranges. For current signals, the accuracy was 91.29% for 0.25s samples and 97.5% across torque ranges.

A High-Precision Aeroengine Bearing Fault Diagnosis Based on Spatial Enhancement Convolution and Vision Transformer

A High-Precision Aeroengine Bearing Fault Diagnosis Based on Spatial Enhancement Convolution and Vision Transformer

MT-ConvFormer: A Multitask Bearing Fault Diagnosis Method Using a Combination of CNN and Transformer

MT-ConvFormer: A Multitask Bearing Fault Diagnosis Method Using a Combination of CNN and Transformer

A Digital Twin System for Two-Stage PMSM Rotor and Bearing Faults Identification Based on Deep Learning and Improved-RGB Acoustic Image

A Digital Twin System for Two-Stage PMSM Rotor and Bearing Faults Identification Based on Deep Learning and Improved-RGB Acoustic Image

Causal Discovery for Rolling Bearing Fault Under Missing Data: From the Perspective of Causal Effect and Information Flow

Causal Discovery for Rolling Bearing Fault Under Missing Data: From the Perspective of Causal Effect and Information Flow

Rolling Bearing Fault Diagnosis Based on MTF Encoding and CBAM-LCNN Mechanism

Rolling Bearing Fault Diagnosis Based on MTF Encoding and CBAM-LCNN Mechanism

A Novel Bearing Fault Data Generation Strategy Combining Physical Modeling and CycleGAN Variant for Fault Diagnosis without Real Samples

A Novel Bearing Fault Data Generation Strategy Combining Physical Modeling and CycleGAN Variant for Fault Diagnosis without Real Samples

Artificial Feature Bias Rectified by Self-Supervised Learning for Rolling Bearings Fault Diagnosis Under Limited Labeled Vibration Signals

Artificial Feature Bias Rectified by Self-Supervised Learning for Rolling Bearings Fault Diagnosis Under Limited Labeled Vibration Signals

Data Generation Approach Based on Data Model Fusion: An Application for Rolling Bearings Fault Diagnosis With Small Samples

Data Generation Approach Based on Data Model Fusion: An Application for Rolling Bearings Fault Diagnosis With Small Samples

Research on Bearing Fault Diagnosis Based on Optimized CNN and Information Fusion

Aiming at the problem of the low recognition rate of a bearing fault under a single sensor condition, an intelligent detection method of a bearing fault based on an optimized convolutional neural network (CNN) and information fusion is proposed. Firstly, a NU216 bearing is selected as the research object to prefabricate fault defects. Then, the orthogonal test method is used to design the test scheme, and the vibration signal and acoustic emission signal of the NU216 bearing are collected; Secondly, the mutual feature fusion method of convolutional neural network is used to fuse the vibration signal and the acoustic emission signal. Then, the fused one-dimensional data is converted into a two-dimensional image dataset by continuous wavelet transform. Finally, the three datasets are divided into training sets and test sets, and input into the optimized convolution neural network model for training and testing. The test results show that the accuracy of fault diagnosis based on vibration signal is 95.76%, the accuracy of fault diagnosis based on acoustic emission signal is 92.33%, and the accuracy of fault diagnosis based on fusion signal is 98.59%.

Mechanical compressive properties of TPU 3D printed with various parameters

Shape memory polymer (SMP) is a material that has the ability to recover its original shape from a temporary (deformed) shape by applying external stimuli. The smart scaffold based on SMP is used to enhance delivery, load bearing, and tissue defect filling. Therefore, specimens with the structure of the face-centered cubic were produced under various printing conditions to characterize their effects on the mechanical properties. Fused deposition modeling is utilized to construct the specimens of shape memory thermoplastic polyurethane (MM-3520). Printing parameters with different levels were used in specimen fabrication, including layer thicknesses of 0.1, 0.2, and 0.3 mm, printing temperatures of 210, 220, and 230 ° C, and printing speeds of 20, 30, and 40 mm/sec. We performed the microstructural analysis under a microscope to examine the impact of printing factors on lattice structures. Then there is the compression test, which evaluates mechanical properties such as linear elastic stiffness, collapse stress, plateau stiffness, and densification stress. Analyzing the microstructure of the printed specimens exhibits that the specimens with the highest printing temperature, the lowest printing speed, and a thinner printing layer have better layer adhesion and lower porosities. As well, figures and main effect plots revealed that the specimens printed with a layer height of 0.1mm, a printing temperature of 230 ° C, and a printing speed of 20 mm/s had compressive strengths of 0.6129±0.062, 0.6018±0.106, and 0.6082±0.078 MPa, respectively. These are the highest results in terms of strength compared to other levels of parameters.

A novel noise reduction algorithm and its application in failure information extraction of bearing

The performance of rolling bearing can directly affect the reliability of a rotating machine. Due to the influence of noise and irrelevant components, the fault information is rather weak in a bearing fault, which adds to the difficulty in judgment of bearing state and identification of fault types. To solve the problem of heterogeneous feature information caused by noise jamming in rolling bearing failure, difficulty in extracting fault feature, and misdiagnosis and misjudgment, the paper has proposed a denoising algorithm according to the covariance matrix of a Hankel matrix and information entropy (Hankel-Cov-IE). First, covariance matrix of a Hankel matrix of signals is constructed, and fault signals are preliminarily denoised based on denoising capacity of the covariance matrix and its nature of combining global and local features. To address the selection of sensitive fault components, information entropy (IE) is taken advantage of to screen out the vectors containing abundant feature information from the covariance matrix. The filtered feature vectors are reconstructed in the way of equal weight and faults of rolling bearing are identified with the spectrum of signals reconstructed. The presented algorithm is compared with different typical methods. The results indicate that the presented algorithm can reduce noise more effectively and achieve extraction of bearing failure features more comprehensively and correctly, with correct judgment of fault types. The presented algorithm can correctly extract the feature frequency of bearing fault and recognize the types of faults, and is insensitive to sensor sites, speeds, and failure types of bearings.

Sample Augmentation Using Enhanced Auxiliary Classifier Generative Adversarial Network by Transformer for Railway Freight Train Wheelset Bearing Fault Diagnosis.

Diagnosing faults in wheelset bearings is critical for train safety. The main challenge is that only a limited amount of fault sample data can be obtained during high-speed train operations. This scarcity of samples impacts the training and accuracy of deep learning models for wheelset bearing fault diagnosis. Studies show that the Auxiliary Classifier Generative Adversarial Network (ACGAN) demonstrates promising performance in addressing this issue. However, existing ACGAN models have drawbacks such as complexity, high computational expenses, mode collapse, and vanishing gradients. Aiming to address these issues, this paper presents the Transformer and Auxiliary Classifier Generative Adversarial Network (TACGAN), which increases the diversity, complexity and entropy of generated samples, and maximizes the entropy of the generated samples. The transformer network replaces traditional convolutional neural networks (CNNs), avoiding iterative and convolutional structures, thereby reducing computational expenses. Moreover, an independent classifier is integrated to prevent the coupling problem, where the discriminator is simultaneously identified and classified in the ACGAN. Finally, the Wasserstein distance is employed in the loss function to mitigate mode collapse and vanishing gradients. Experimental results using the train wheelset bearing datasets demonstrate the accuracy and effectiveness of the TACGAN.

Early Detection of Ball Bearing Faults Using the Decision Tree Method

Bearings are one of the important components in the machine that functions as a holder and positions the shaft alignment radially when rotating. Statistics show that about 50% of failures in electric motors are related to bearings. Therefore, monitoring bearing performance and efficiency before damage occurs is necessary to avoid more serious damage and save repair costs. This research aims to build a classification model that can identify bearings in normal condition and 6 types of damage (inner crack, outer crack, ball crack, and a combination of both) using the HUST dataset. The model building process begins with collecting datasets, processing and extracting dataset features, building classification models and evaluating the models that have been made. A decision tree is a type of supervised machine learning used to categorize or make predictions based on how a previous set of questions were answered. The model is a form of supervised learning, meaning that the model is trained and tested on a set of data that contains the desired categorization. The results of the decision tree model that has been built are able to identify bearing damage with an accuracy of 94.47%.

A Lightweight and Small Sample Bearing Fault Diagnosis Algorithm Based on Probabilistic Decoupling Knowledge Distillation and Meta-Learning.

Rolling bearings play a crucial role in industrial equipment, and their failure is highly likely to cause a series of serious consequences. Traditional deep learning-based bearing fault diagnosis algorithms rely on large amounts of training data; training and inference processes consume significant computational resources. Thus, developing a lightweight and suitable fault diagnosis algorithm for small samples is particularly crucial. In this paper, we propose a bearing fault diagnosis algorithm based on probabilistic decoupling knowledge distillation and meta-learning (MIX-MPDKD). This algorithm is lightweight and deployable, performing well in small sample scenarios and effectively solving the deployment problem of large networks in resource-constrained environments. Firstly, our model utilizes the Model-Agnostic Meta-Learning algorithm to initialize the parameters of the teacher model and conduct efficient training. Subsequently, by employing the proposed probability-based decoupled knowledge distillation approach, the outstanding performance of the teacher model was imparted to the student model, enabling the student model to converge rapidly in the context of a small sample size. Finally, the Paderborn University dataset was used for meta-training, while the bearing dataset from Case Western Reserve University, along with our laboratory dataset, was used to validate the results. The experimental results demonstrate that the algorithm achieved satisfactory accuracy performance.

Comparative study of kurtosis and L-kurtosis for bearing fault classification in induction motors

This study investigates the effectiveness of L-kurtosis as a robust alternative to traditional kurtosis for identifying and categorizing rolling bearing faults in vibration signals. By comparing L-kurtosis-energy and kurtosis-energy features derived from wavelet packet decomposition (WPD) coefficients; this research evaluates their performance using a multi-layer perceptron neural network (MLP-NN). Experimental data encompassing various rotating speeds, fault types, and severities were utilized to train and test the MLP-NN on both healthy and defective bearing conditions. The results demonstrate that while kurtosis-energy achieved 95.63% accuracy in defect classification, replacing kurtosis with L-kurtosis significantly enhanced accuracy to 99.92%. This improvement underscores the resilience of L-kurtosis to outliers and its ability to handle non-normally distributed vibration signals effectively. The findings affirm the potential of L-kurtosis-energy features to improve fault detection methodologies, making them more reliable for industrial applications. This study highlights the importance of robust diagnostic tools for advancing predictive maintenance strategies and ensuring operational reliability.

Step-Wise Parameter Adaptive FMD Incorporating Clustering Algorithm in Rolling Bearing Compound Fault Diagnosis

Ideally, the vibration signal of a rolling bearing should be symmetrical. However, in practical operation, the vibration signals in both time and frequency domains often exhibit asymmetry due to factors such as load, speed, and wear. The relatively weak composite fault characteristics are easily masked. Although the Feature Modal Decomposition (FMD) method is outstanding in diagnosing composite faults in bearings, its effectiveness is easily constrained by parameter selection. To address this, this paper proposes a stepwise parameter adaptive FMD method combined with a clustering algorithm, specifically designed for diagnosing composite faults in rolling bearings. Firstly, this study employs the Density Peak Clustering algorithm to determine the number of modes n in the composite fault vibration signal. Subsequently, considering the signal spectral energy and modal characteristics, a new composite fault index is formulated, namely, the adaptive weighted frequency domain kurtosis-to-information entropy ratio, as the fitness function. The Whale Optimization Algorithm determines the filter length L and the number of segments K, thereby achieving step-wise signal decomposition. Through in-depth analysis of signal symmetry and asymmetry, simulation and experimental verification confirm the effectiveness of this method. Compared with four other index-optimized FMD methods and traditional techniques, this method significantly reduces the influence of parameters on FMD, is capable of separating the characteristic frequencies related to composite faults, and performs excellently in the diagnosis of composite faults in rolling bearings.