A rolling bearing fault diagnosis method based on the improved sparrow search algorithm optimized VMD and multi-scale convolutional neural networks

Research on rolling bearing fault diagnosis method based on AMVMD and convolutional neural networks

To achieve the goal of automated rolling bearing fault diagnosis, a variational mode decomposition (VMD) based diagnosis scheme was proposed. VMD was firstly used to decompose the vibration signals into a series of band-limited intrinsic mode functions (BLIMFs). Subsequently, the multiscale fractal dimension (MSFD) and multiscale energy (MSEN) of each BLIMF were calculated and combined together as features of the original vibration signals. In an attempt to accelerate the classification speed, one-way analysis of variance (ANOVA) test was adopted to extract significant features from the redundant features. Finally, those significant features were fed into the optimized support vector machine (SVM), which was optimized by the genetic algorithm (GA), for classification. Experimental results on the international public Case Western Reserve University bearing data indicate the effectiveness of the proposed method with a classification accuracy of 99.75 % for seven classes. Moreover, our approach also shows good anti-noise performance in different signal-to-noise ratios (SNRs).

To address the issues of inadequate fault feature extraction and insufficient diagnostic model accuracy in rolling bearings, this paper proposes a fault diagnosis model (SFOA-VMD-AMSCNN). This model integrates Variational Modal Decomposition (VMD), optimised via the Starfish Optimisation Algorithm (SFOA), with a Multi-Scale Convolutional Neural Network incorporating a Fusion Attention Mechanism. First, to resolve the modal aliasing issue arising from challenging parameter selection in Variational Modal Decomposition (VMD), the Starfish Optimisation Algorithm (SFOA) is employed to adaptively optimise the number of modal decompositions and the penalty factor. This establishes the SFOA-VMD method, enabling precise decomposition and feature extraction from bearing vibration signals. Building upon this foundation, a novel attention mechanism multi-scale convolutional neural network is constructed. This network employs three parallel branches, utilising distinct convolutional kernels to extract both local and long-range dependent features from the signal. It incorporates a SE attention mechanism to enable adaptive recalibration of feature channels, thereby enhancing the expressive power of discriminative features. Experimental results on bearing data from West China Medical College demonstrate that this method achieves an average accuracy of 99.94% across ten fault classifications, validating the model’s exceptional diagnostic performance and generalisation capability.

In response to the problem of difficulty in extracting fault features of rolling bearings in wind turbine transmission systems under complex working conditions, which limits the accuracy of fault diagnosis. This article proposes an Adaptive Deep Learning based Rolling Bearing Fault Diagnosis Method (ADLM). Introducing dynamic convolution into Convolutional Neural Networks (CNNs) can adaptively capture data features; At the same time, the fishing optimization algorithm (CFOA) was used to optimize the hyperparameters of the bidirectional long short-term memory network (BiLSTM), and the CFOA-BiLSTM network was constructed to fully leverage its advantages in time series analysis. The specific implementation steps are as follows: first, preprocess the collected vibration signals and divide the processed dataset into a training set and a testing set; Then, parallel adaptive convolutional neural networks (ACNN) are used to process the training set and extract spatial domain local features from the vibration signal; Then, the features extracted from the two branches are weighted and fused through a dynamic weight adjustment mechanism, and the fused features are input into the CFOA-BiLSTM network to further capture the time-dependent features of the signal; Finally, the extracted features are input into the classifier to complete model training, and the model performance is evaluated using a test set. Experimental verification shows that on the dataset of Southeast University, the diagnostic accuracy of the ADLM model reached 98.52 %, demonstrating good reliability, robustness, and superiority in the diagnosis of rolling bearing faults.

To enhance the precision of rolling bearing fault diagnosis, an intelligent hybrid approach is proposed in this paper for signal processing and fault diagnosis in small samples. This approach is based on advanced techniques, combining parameter optimization variational mode decomposition weighted by multiscale permutation entropy (MPE) with maximal information coefficient and multi-target attention convolutional neural networks (MTACNN). First, an improved variational mode decomposition (VMD) is developed to denoise the raw signal. The whale optimization algorithm was used to optimize the penalty factor and mode component number in the VMD algorithm to obtain several intrinsic mode functions (IMFs). Second, separate MPE calculations are performed for both the raw signal and each of the IMF components obtained from the VMD decomposition; the results are used to calculate the maximum information coefficient (MIC). Subsequently, each MIC is normalized and converted to a weight coefficient for signal reconstruction. Ultimately, the reconstructed signals serve as input to the MTACNN for diagnosing rolling bearing faults. Results demonstrate that the signal processing approach exhibits superior noise reduction capability through simple processing. Furthermore, compared to several similar approaches, The method proposed for fault diagnosis achieves superior performance levels in the fault pattern recognition target and the fault severity recognition target.

For the hassle that the fault of the rolling bearing is challenging to identify, a method for identifying the damage degree of rolling bearing based on the combination of Variational Mode Decomposition (VMD) and Gath-Geva (GG) fuzzy clustering is proposed. The method decomposes the known rolling bearing fault signal by VMD, determines the number of decomposed modes based on the magnitude of the component frequency center, and forms the obtained eigenmode components into an initial eigenmatrix for singular value decomposition. The three largest singular values are selected as the input of the GG clustering algorithm, and the membership matrix and cluster center of the known fault signals are obtained. The damage degree of the rolling bearing is identified by the Hamming closeness between the initial membership matrix of the signal to be measured and the cluster centers with clear fault signal. Rolling bearings vibration data verify the effectiveness of the method. The experimental results show that the method in this paper can accurately extract the fault information contained in the vibration signal, realize the correct fault type identification, and provide a reliable basis for the rolling bearing fault diagnosis method.

To address the challenges of weak fault features and strong non-stationarity in early-stage vibration signals, this study proposes a novel fault diagnosis method combining enhanced variational mode decomposition (VMD) with a structurally improved GoogLeNet. Specifically, an improved wild horse optimizer (IWHO) with tent chaotic mapping is employed to automatically optimize critical VMD parameters, including the number of modes K and the penalty factor α, enabling precise decomposition of non-stationary signals to extract weak fault features. The vibration signal is decomposed, and the top five intrinsic mode functions (IMFs) are selected based on the kurtosis criterion. Time–frequency features are then extracted from these IMFs and input into a modified GoogLeNet classifier. The GoogLeNet structure is improved by replacing standard n × n convolution kernels with cascaded 1 × n and n × 1 kernels, and by substituting the ReLU activation function with a parameterized TReLU function to enhance adaptability and convergence. Experimental results on two public rolling bearing datasets demonstrate that the proposed method effectively handles non-stationary signals, achieving 99.17% accuracy across four fault types and maintaining over 95.80% accuracy under noisy conditions.

Rolling bearings are vital components in rotary machines where even minor faults can lead to the collapse of the entire mechanical system, resulting in significant economic losses. The condition monitoring and fault diagnosis of rolling bearings are crucial for safe operation of complex industrial systems. Traditional fault diagnosis methods mainly depend on expert experience, which leads to obvious deficiencies in diagnostic efficiency, especially in terms of accuracy. To address the issues of traditional bearing fault diagnosis relying on expert experience and low efficiency and accuracy of fault diagnosis, a bearing fault diagnosis model combining synchrosqueezed wavelet transform (SWT) and SE-ResNet is proposed. Firstly, the one-dimensional non-stationary bearing vibration signal is converted into a high-frequency two-dimensional time-frequency image through SWT, which serves as the input for the convolutional neural network; Secondly, an attention mechanism is introduced to construct an SE-ResNet model for extracting two-dimensional image features, which enhances the attention of important features and thus enhances the network's representation ability; Finally, the results of fault diagnosis are output through Softmax layer. To validate the capacity of the proposed model, experiments with the bearing dataset collected from Case Western Reserve University (CWRU) were conducted. The results show that the transfer learning accuracy of the CWRU dataset is 86.25% to 99%. This method has good feasibility, good convergence performance, and has high practical application value.

Learning multiscale and deep representations for classifying remotely sensed imagery

Aiming to address the suboptimal diagnostic efficacy of conventional approaches in rolling bearing fault diagnosis under complex operational scenarios—particularly when subjected to intense noise contamination—this study introduces a novel fault diagnosis framework leveraging the synergistic attention mechanisms of MSCNN and GRU.The workflow begins with conditioning vibration signals sourced from the CRUW bearing dataset, followed by robust extraction of multi-scale fault features through a multi-scale convolutional neural network (MSCNN). To enhance temporal modeling, a gated recurrent unit (GRU) is integrated to extract temporal dependencies from the signals. Further improvements are achieved by enhancing the base model with Transformer layers and a convolutional attention module (CBAM), respectively. Experimental validation is conducted to compare the diagnostic performance of three configurations MSCNN+GRU, MSCNN+GRU+Transformer, and MSCNN+GRU+CBAM—under varying noise conditions. Outcomes demonstrate that the MSCNN+GRU+CBAM model exhibits superior fault detection accuracy and optimal generalization capacity across different load and rotational speed settings, confirming its notable advantage in rolling bearing fault diagnosis.

With the automation and complexity of industry, the production planning and scheduling of the smart factory depends on the proper functioning of each part. As a key component of rotating machinery, the faults of rolling bearing is prone to effects under complex and changing working conditions and coupled with other components. Therefore, it is necessary to carry out efficient and accurate fault diagnosis for rolling bearings to ensure the reliability and stability of mechanical equipment operation. This paper introduces a multi-scale rolling bearing fault diagnosis method (MSF-CNN) based on dilated convolution which is able to automatically capture complementary data and feature information of different scales of the original vibration signal through multi-scale convolution to achieve end-to-end fault diagnosis. The experimental results show that the proposed fault diagnosis method achieves an average accuracy of 99.35% under three load conditions. In addition, this method can provide the cross domain data solution to reduce the dependence on missing label data, which achieves the fault diagnosis accuracy of 89+%. It verifies the validity of our proposal.

We propose Multi-scale convolutional recurrent neural networks (MCRNN) and data augmentation methods to detect polyphonic sound event with few training data. MCRNN consists of Multi-scale convolutional neural networks (MCNN) and recurrent neural networks (RNN). MCNN concatenates the higher level features extracted using multiple convolution kernels with different scales from the time domain and frequency domain at the same time. RNN is able to capture the longer term temporal context characteristics. A novel background spectrum random replacement (BSRR) data augmentation method is applied to expand training data, which uses standard normal distribution data with randomly selected position and length instead of the original time-domain, frequency-domain or time-frequency domain background spectrum features. Our method is tested on the datasets of DCASE 2019 Task3 (T3). The experimental results showed that the MCRNN and BSRR data augmentation method are efficient. We achieved better results than the first place and the single advanced on the T3 by applying BSRR and SpecAugment data augmentation method simultaneously. On the evaluation dataset (T3-eval), our best result shows 0.05 and 0.975 of error rate (ER) and F1 respectively. Our method got the best performance and relatively improved 17% and 1% than the corresponding values of the single advanced.

In order to realize the quantitative assessment of muscle strength in hand function rehabilitation and then formulate scientific and effective rehabilitation training strategies, this paper constructs a multi-scale convolutional neural network (MSCNN) - convolutional block attention module (CBAM) - bidirectional long short-term memory network (BiLSTM) muscle strength prediction model to fully explore the spatial and temporal features of the data and simultaneously suppress useless features, and finally achieve the improvement of the accuracy of the muscle strength prediction model. To verify the effectiveness of the model proposed in this paper, the model in this paper is compared with traditional models such as support vector machine (SVM), random forest (RF), convolutional neural network (CNN), CNN - squeeze excitation network (SENet), MSCNN-CBAM and MSCNN-BiLSTM, and the effect of muscle strength prediction by each model is investigated when the hand force application changes from 40% of the maximum voluntary contraction force (MVC) to 60% of the MVC. The research results show that as the hand force application increases, the effect of the muscle strength prediction model becomes worse. Then the ablation experiment is used to analyze the influence degree of each module on the muscle strength prediction result, and it is found that the CBAM module plays a key role in the model. Therefore, by using the model in this article, the accuracy of muscle strength prediction can be effectively improved, and the characteristics and laws of hand muscle activities can be deeply understood, providing assistance for further exploring the mechanism of hand functions .

Deep learning methods are widely used in the field of rolling bearing fault diagnosis and produce good results when faced with datasets with roughly equal numbers of normal and faulty samples. However, real-world data often has a serious imbalance, with the number of fault samples being significantly less than the number of normal samples. This dataset imbalance challenges the performance of traditional deep learning methods. To address this problem, this paper proposes an efficient imbalanced data rolling bearing fault diagnosis method. The method consists of two parts: a deep learning network based on a multi-scale self-attention mechanism and a novel loss function. In terms of the deep learning network, firstly, the one-dimensional vibration signal is converted into a two-dimensional image through the Gramian angular field. This conversion maximises the inherent feature extraction capability of the network. Subsequently, the multi-scale learning capability of the network is enhanced by implementing different expansion rates for the head of the multi-scale self-attention mechanism. This nuanced approach allows the network to capture the underlying information more efficiently. Finally, the inclusion of Ghost bottlenecks and feature pyramid networks (FPNs) helps to optimise network efficiency and improve generalisation performance. A novel loss function is also proposed to make the method more suitable for imbalanced data. During the training process, the classification of samples in each class is analysed using the recall metric of imbalanced classification and the real-time recall is used as a weight to weaken the dominance of the majority class. This weighting ensures the adaptability of the method to imbalanced datasets. The proposed method is evaluated using rolling bearing datasets from Case Western Reserve University, USA, and Southeast University, China. Comparison results with other state-of-the-art deep learning methods show that the proposed method has a robust performance when dealing with imbalanced data.

The bearing fault diagnosis is of vital significance in maintaining the safety of rotation machine. Among various fault detection techniques, the diagnosis based on vibration signal is widely applied in monitoring the condition of rotation machine. Variational mode decomposition (VMD) is a novel signal analysis method, which can decompose a multi-component signal into a certain number of band-limited intrinsic mode functions (BLIMFs) nonrecursively. VMD could overcome some problems such as mode mixing, the inference of noise, the determination of wavelet base, which exist in empirical mode decomposition, ensemble empirical mode decomposition, wavelet transform, respectively. However, the empirical selection of the parameters for VMD would affect the result of the decomposition. This paper presents an adaptive VMD method with parameter optimization for detecting the localized faults of rolling bearing. Kurtosis, sensitive to transient impulsive components, is employed as optimization index to evaluate the performance of the VMD. Two parameters in the VMD, namely the number of decomposition modes and data-fidelity constraint, are optimized synchronously based on the kurtosis index through artificial fish swarm algorithm. Executing VMD with the acquired parameters, the optimal BLIMF is obtained. The spectrum analysis of the optimal BLIMF could identify the characteristic frequency caused by the localized crack effectually. The validity of the proposed method is proved by means of a cyclic transient impulse response signal and two experiments with practical vibration signals of rolling bearings. Compared to several existing methods, the proposed method demonstrates reinforced results.