Neural Network Fault Diagnosis Research Articles

Information-based Gradient enhanced Causal Learning Graph Neural Network for fault diagnosis of complex industrial processes

By representing the embedded components and their interactions in industrial systems as nodes and edges in a graph, Graph Neural Networks (GNNs) have achieved outstanding results due to their ability to model statistical correlations. However, these correlations may not capture the true causal relationships within the data, thereby impairing the model’s performance in fault diagnosis.To address this issue, an Information-based Gradient enhanced Causal Learning Graph Neural Network (IGCL-GNN) is proposed for fault diagnosis of complex industrial processes. First, the information theory in graph representations is theoretically analyzed and the optimization objectives are derived separately for the causal and non-causal parts of the graph neural network, which decouple it into a multi-objective optimization problem. Then, to optimize such problem, a causal disentanglement layer is designed in the graph network that could effectively separate causal and non-causal information in graph representations. Thirdly, a novel gradient reactivation method is proposed to dynamically filter features from the disentangled layers, thereby capture the causal representations of graph data more accurately. For robust and efficient optimization, the multi-objective gradient descent algorithm is employed in this paper. Finally, comparative experiments were conducted on the three-phase flow facility (TFF) dataset, achieving a fault diagnosis accuracy of 98.42% for our proposed method.

Causal intervention graph neural network for fault diagnosis of complex industrial processes

With the development of industry and manufacturing, the mechanical structures of equipment have become intricate and complex. Due to the interaction between components, once a failure occurs, it will propagate through the industrial processes, resulting in multiple sensor anomalies. Identifying the root causes of faults and eliminating interference from irrelevant sensor signals are critical issues in enhancing the stability and reliability of intelligent fault diagnosis. The components of industrial processes and their interactions can be represented by a structural attribute graph. The causal subgraph formed by fault signals determines the fault mode, while irrelevant sensor signals constitute a non-causal subgraph. The structure of non-causal subgraphs is relatively simple, and graph neural networks tend to use this part as a shortcut for prediction, leading to a significant decrease in prediction accuracy. To address this issue, a causal intervention graph neural network (CIGNN) framework is proposed. First, the sensor signals are constructed into structural attribute graphs using an attention mechanism. Due to causal and confounding features are highly coupled in graphs, explicitly decoupling them is almost impossible. Then, we design an instrumental variable to implement causal intervention to mitigate the confounding effect. Experimental results on two complex industrial datasets demonstrate the reliability and effectiveness of the proposed method in fault diagnosis.

Research on bearing fault diagnosis based on improved genetic algorithm and BP neural network

Health monitoring and fault diagnosis of rolling bearings are crucial for the continuous and effective operation of mechanical equipment. In order to improve the accuracy of BP neural network in fault diagnosis of rolling bearings, a feature model is established from the vibration signals of rolling bearings, and an improved genetic algorithm is used to optimize the initial weights, biases, and hyperparameters of the BP neural network. This overcomes the shortcomings of BP neural network, such as being prone to local minima, slow convergence speed, and sample dependence. The improved genetic algorithm fully considers the degree of concentration and dispersion of population fitness in genetic algorithms, and adaptively adjusts the crossover and mutation probabilities of genetic algorithms in a non-linear manner. At the same time, in order to accelerate the optimization efficiency of the selection operator, the elite retention strategy is combined with the hierarchical proportional selection operation. Using the rolling bearing dataset from Case Western Reserve University in the United States as experimental data, the proposed algorithm was used for simulation and prediction. The experimental results show that compared with the other seven models, the proposed IGA-BPNN exhibit superior performance in both convergence speed and predictive performance.

Reinforced convolutional neural network fault diagnosis of industrial production systems

A new fault diagnosis method is proposed combining multiple models into a multiscale residual jagged dilated convolutional neural network with long short-term memory (MRJDCNN-LSTM). This method has strong feature extraction ability and can effectively extract and classify fault features. First, a new multi-scale residual jagged dilated convolution neural network (MRJDCNN) model is designed by incorporating the principle of residual learning into the improved jagged dilated convolution, and then it is combined with the long short term memory(LSTM) network delicately. This not only greatly improves the extraction rate of nonlinear high-latitude spatial features at different scales while preventing model degradation, but also subtly extracts time-related features, effectively reducing the loss of necessary features, and thus greatly improving the accuracy of model diagnosis. The simulation and comparative experiments of Tennessee-Eastman (TE) process and industrial coking furnace both prove that the constructed model has excellent performance.

Enhancing interpretability in neural networks for nuclear power plant fault diagnosis: A comprehensive analysis and improvement approach

Deep neural networks, as applied in the field of nuclear power fault diagnosis, have garnered significant attention alongside advancements in artificial intelligence technology. However, the “black box” nature of deep learning models has raised concerns regarding their deployment in scenarios demanding high safety standards, such as nuclear power plants. In this paper, we innovatively propose the utilization of an explainable artificial intelligence method, grounded in game theory, to conduct a detailed analysis of the diagnostic behavior of neural network models applied to nuclear power plants. By leveraging SHAP, the decision-making processes of these opaque models are demystified, offering insights into how and why they arrive at their predictions. In this study, two of the most widely utilized neural network frameworks are applied across six representative fault diagnosis cases for a comprehensive analysis. Based on its analysis results, a novel SHAP-Enhanced Feature Selection for Efficient Neural Network Fault Diagnosis strategy is proposed, significantly reducing model complexity without sacrificing diagnostic performance. This research breaks through the limitations of data-driven models used in the field of nuclear power fault diagnosis, conducts an interpretable analysis of their behavior, and proposes an improved strategy based on the analysis results, contributing to the practical application of data-driven models in nuclear power.

Fault diagnosis of drone motors driven by current signal data with few samples

Multi rotor unmanned aerial vehicles (UAVs) are extensively utilized across various domains, and the motor constitutes a pivotal element in the UAV power system. The majority of UAV failures and crashes stem from motor malfunctions, underscoring the imperative need for comprehensive research on fault diagnosis in UAV motors to ensure the stable and reliable execution of flight tasks. This study focuses on quadrotor UAVs as the research subject and devises targeted fault simulation experiments based on the structural features and operational characteristics of the DC brushless motor used in quadrotor UAVs, specifically examining the stator, rotor, and bearings. To address challenges related to the UAV’s own loads, limited space for redundant parts, and the high cost and difficulty associated with installing sensors for traditional fault diagnostic signals such as vibration and temperature, this study opts to use current signals as a substitute. This approach resolves the issue of challenging data collection for UAVs and investigates a current signal based fault diagnosis method for UAV motors. Lastly, in response to the limited training samples available for fault data due to the UAV’s highly sensitive characteristics regarding the health status of its components and flight stability, traditional machine learning and deep learning methods encounter difficulties in identifying representative features with a small number of training samples, leading to the risk of overfitting and reduced model accuracy in fault diagnosis. To overcome this challenge, we propose a hybrid neural network fault diagnosis model that incorporates a width learning system and a convolutional neural network (CNN). The width learning system eliminates temporal characteristics from the original current signal, capturing more comprehensive and representative sample features in the width feature space. Subsequently, the CNN is employed for feature extraction and classification tasks. In empirical small sample fault diagnosis experiments using current signal data for UAV motors, our proposed model outperforms other models used for comparison.

Application of continuous wavelet transform and convolutional neural networks in fault diagnosis of PMSM stator windings

Efficiency, reliability, and durability play a key role in modern drive systems in line with the Industry 4.0 paradigm and the sustainability trend. To ensure this, highly efficient motors and appropriate systems must be deployed to monitor their condition and diagnose faults during the operation. For these reasons, in recent years, more and more research has been focused on developing new methods for fault diagnosis of permanent magnet synchronous motors (PMSMs). This paper proposes a novel hybrid method for the automatic detection and classification of PMSM stator winding faults based on combining the continuous wavelet transform (CWT) analysis of the negative sequence component of the stator phase currents with a convolutional neural network (CNN). CWT scalogram images are used as the inputs of the CNN-based interturn short circuits fault classifier model. Experimental tests were carried out to verify the effectiveness of the proposed approach under various motor operating conditions and at an incipient stage of fault propagation. In addition, the effects of the input image format, CNN structure, and training process parameters on model accuracy and classification effectiveness were investigated. The results of the experimental tests confirmed the high effectiveness of fault detection (99.4%) and classification (97.5%), as well as other important advantages of the developed method.

A new fault diagnosis of rolling bearing on FFT image coding and L-CNN

To address the problems of low diagnostic accuracy and slow diagnostic speed of the convolutional neural network (CNN) fault diagnosis method in rolling bearing diagnosis, a new rolling bearing fault diagnosis method based on fast Fourier transform (FFT) image coding and lightweight-CNN (L-CNN) is proposed. The method is mainly divided into three stages: firstly, the original signal is reconstructed by noise reduction using a joint noise reduction method of complete ensemble empirical mode decomposition with adaptive noise, permutation entropy, and wavelet threshold denoise; then, the frequency spectra and phase spectra feature fusion data of the noise-reduced and reconstructed bearing vibration signals are obtained by FFT, the feature fusion data are encoded into a heat map, and the image coding data-set is fed into an improved L-CNN for fault diagnosis. Experiments were carried out using the Guangdong University of Petrochemical Technology bearing fault data-set and the Case Western Reserve University bearing fault data-set with diagnostic accuracies of 98.75% and 99%, respectively. The results demonstrate that the method can effectively classify bearing fault vibration signals with the advantages of a fast diagnosis, high accuracy, and good generalization ability.

An Improved Fault Localization Method for Direct Current Filters in HVDC Systems: Development and Application of the DRNCNN Model

This paper focus on direct current (DC) filter grounding faults to propose a novel dilated normalized residual convolutional neural network (DRNCNN) fault diagnosis model for high-voltage direct current (HVDC) transmission systems. To address the insufficiency of the traditional model’s receptive field in dealing with high-dimensional and nonlinear data, this paper incorporates dilated convolution and batch normalization (BN), significantly enhancing the CNN’s capability to capture complex spatial features. Furthermore, this paper integrates residual connections and parameter rectified linear units (PReLU) to optimize gradient propagation and mitigate the issue of gradient vanishing during training. These innovative improvements, embodied in the DRNCNN model, substantially increase the accuracy of fault detection, achieving a diagnostic accuracy rate of 99.28%.

Incremental learning with multi-fidelity information fusion for digital twin-driven bearing fault diagnosis

Digital twin (DT)-driven intelligent fault diagnosis (IFD) has been a hot topic, which can support personalized monitoring of critical machinery. A central challenge is that diagnostic models using deep learning (DL) suffer from the problem of catastrophic forgetting if personalized faults occur in dynamic environments. To deal with this issue, this article presents a class-incremental learning method with multi-fidelity information fusion (MFIF-CIL) for the continuous diagnosis of key faults in rolling bearings. First, an effective bearing DT model is constructed to generate enough low-fidelity (LF) simulation data. Second, feature boosting is developed to fit the residuals with distribution drifts between old classes and new classes, which helps prevent the problem of catastrophic forgetting. Last, the MFIF module is proposed for multi-fidelity knowledge transfer and fusion to leverage LF simulation data to improve the class-incremental learning ability of feature boosting with limited high-fidelity (HF) physical data. The testing datasets consisting of the measured signals are utilized as testing datasets of optimal incremental neural networks for fault diagnosis. The proposed MFIF-CIL-1 (using 15 HF data and 100 LF data as exemplars) and MFIF-CIL-2 (using 20 HF data and 100 LF data as exemplars) obtain the average diagnostic accuracies of 96.87% and 98.10%, respectively. The MFIF-CIL-2 only uses 41.93% of the training time required by the joint training method. These satisfying results demonstrate that the MFIF-CIL can effectively diagnose different health conditions over time and provide a tradeoff between relatively low experimental costs and high accuracy.

Interpretable parallel channel encoding convolutional neural network for bearing fault diagnosis

Interpretability plays a crucial role in the application of neural networks for fault diagnosis. Integrating preprocessing methods into neural networks can enhance interpretability while preserving their ‘end-to-end’ characteristics. However, when only redesigning the first layer, subsequent structures still exhibit limited transparency. Additionally, traditional convolution structure is ill-suited for analyzing the readable feature maps derived from vibration signals. To address these challenges, this paper proposes a novel convolution structure for the parameterized signal processing function of the first-layer convolution kernel. This structure incorporates channel mixing for feature augmentation, designs a condensed feature encoder for aggregating and compressing features channel-by-channel, ensures the interpretability of feature map processing, and obtains condensed feature codes to propose smooth activation layer-wise relevance propagation (SA-LRP) method that to perform interpretability analysis. Additionally, cubic feature screening is implemented for diagnostic classification to improve structural fitness. We design experiments using multiple datasets to test various indicators of the structure. The results confirm that connecting our convolution architecture for subsequent analysis outperforms other convolution architectures for the convolution kernel of the first-layer parameterized signal processing function. The interpretability of the model is evaluated through SA-LRP method and validates the interpretability of the model.

Rotating Machinery Fault Diagnosis with Limited Multisensor Fusion Samples by Fused Attention-Guided Wasserstein GAN

As vital equipment in modern industry, the health state of rotating machinery influences the production process and equipment safety. However, rotating machinery generally operates in a normal state most of the time, which results in limited fault data, thus greatly constraining the performance of intelligent fault diagnosis methods. To solve this problem, this paper proposes a novel fault diagnosis method for rotating machinery with limited multisensor fusion samples based on the fused attention-guided Wasserstein generative adversarial network (WGAN). Firstly, the dimensionality of collected multisensor data is reduced to three channels by principal component analysis, and then the one-dimensional data of each channel are converted into a two-dimensional pixel matrix, of which the RGB images are obtained by fusing the three-channel two-dimensional images. Subsequently, the limited RGB samples are augmented to obtain sufficient samples utilizing the fused attention-guided WGAN combined with the gradient penalty (FAWGAN-GP) method. Lastly, the augmented samples are applied to train a residual convolutional neural network for fault diagnosis. The effectiveness of the proposed method is demonstrated by two case studies. When training samples per class are 50, 35, 25, and 15 on the KAT-bearing dataset, the average classification accuracy is 99.9%, 99.65%, 99.6%, and 98.7%, respectively. Meanwhile, the methods of multisensor fusion and the fused attention mechanism have an average improvement of 1.51% and 1.09%, respectively, by ablation experiments on the WT gearbox dataset.

Fuzzy RBF neural network fault diagnosis method based on knowledge and data fusion for the recoil system

To improve the accuracy of fault diagnosis for recoil systems under multiple operating conditions, a fuzzy RBF neural network (Radial Basis Function, RBF) fault diagnosis method based on knowledge and data fusion is proposed. A kinetic model for the recoil system is first established to describe the system’s behavior. Next, fuzzy RBF neural network is used to establish the relationship between abnormal operating parameters and fault causes, achieving a fault cause diagnosis accurately based on the integration of expert experience knowledge and system operation data. A study case demonstrate that the algorithm has strong knowledge and data fusion capabilities and can effectively identify faults in recoil system.

Causal-Trivial Attention Graph Neural Network for Fault Diagnosis of Complex Industrial Processes

Causal-Trivial Attention Graph Neural Network for Fault Diagnosis of Complex Industrial Processes

Fault prediction of electronic devices based on attention mechanism time-series point process

In recent years, with the development of high-precision sensors, computer technology and artificial intelligence, it makes the data-driven intelligent fault warning and diagnosis method gradually become a research hotspot. Aiming at the low utilization of fault diagnosis information of complex electronic devices and the problems of convergence and training time of traditional neural network fault diagnosis models, a point-in-time process generation model without using intensity function is proposed. The model uses the Wasserstein distance to construct the loss function, which facilitates the measurement of the deviation between the model distribution and the true distribution, and uses a self-concern mechanism to describe the degree of influence of the historical events on the current events, which makes the model interpretable and more capable of generalization. Comparative experiments show that without a priori information about the intensity function, the method reduces the relative error rate by 3.59 %, improves the fault prediction accuracy by 3.91 %, and has a better overall fit than the RNN-like generative model and the great likelihood model. Example analysis shows that the model has good prediction accuracy and provides a feasible solution for real-time fault diagnosis of complex electronic devices.

Optimized convolutional neural networks for fault diagnosis in wastewater treatment processes

An optimized deep learning model with high classification performance was proposed for fault diagnosis in wastewater treatment processes.

Physics-Embedded Recurrent Graph Neural Network for Fault Diagnosis of Complex Systems

Physics-Embedded Recurrent Graph Neural Network for Fault Diagnosis of Complex Systems

Unsupervised rolling bearing fault diagnosis method across working conditions based on multiscale convolutional neural network

In practical engineering, the features of rolling bearing vibration signals often vary in distribution under different working conditions, and obtaining sample labels for target working conditions can be challenging. Therefore, a multi-scale depth subdomain adaptive convolutional neural network (MS-DSACNN) fault diagnosis method is proposed. The MS-DSACNN method is based on a multi-scale feature extraction method combined with an improved 1D-ConvNeXt architecture, which fuses low- and high-level features into multi-scale fault features. A channel attention mechanism is also introduced to adaptively assign weights to the fused multi-scale features. The local maximum mean discrepancy is constructed by extracting features and their corresponding source and pseudo-label inputs for the source and target working conditions, respectively, to construct the loss function for backpropagation. The MS-DSACNN model was evaluated in experiments with two different loads and one variable speed dataset. The experimental results indicate that MS-DSACNN has higher feature extraction capability and diagnostic accuracy than other unsupervised cross-condition fault diagnosis techniques. Additionally, the scenario where the signal captured by the sensor contains a considerable amount of noise in actual working conditions is considered.

Application of WCA-RBF Neural Network in Fault Diagnosis of Analog Circuits

With the fast growth of electronic technology today, the hybrid circuit of analog and digital circuits has become a trend in the growth of electronic technology. To address fault diagnosis in analog circuits, Haar wavelet is applied for fault feature extraction. The K-means clustering method and pseudoinverse algorithm were used to optimize the center value and weight value of the radial basis function neural network, respectively. The adaptive step size was improved in the wolf pack algorithm, the parameters of the radial basis function neural network were optimized based on the wolf pack algorithm, and a wolf pack algorithm optimized radial basis function neural network model was constructed. The test results show that this model converges after 40 times of training, with an error value of 10−3 and an average value of the mean squared error of 0.45. Comparing the fault diagnosis rates of the original model, genetic algorithm optimized radial basis function neural network model, and wolf colony algorithm optimized radial basis function neural network model, the last model has the best fault diagnosis rate, reaching 95.52%. The wolf colony algorithm is utilized to optimize the radial basis function neural network model to diagnose the faults in the standard filter circuit, and the fault diagnosis rate reaches 96.17%. The findings express that the radial basis function neural network model optimized by the wolf colony algorithm has a good diagnosis effect for different faults in the analog circuit.

Attention mechanism‐based multisensor data fusion neural network for fault diagnosis of autonomous underwater vehicles

AbstractThe autonomous underwater vehicle (AUV) frequently operates in harsh underwater environments, and timely fault diagnosis of the AUV can prevent mission failure and equipment loss. Data‐driven methods based on a single data source have been widely utilized for fault diagnosis of the AUV because they do not require the construction of complex mechanism models and have high fault diagnosis accuracy. However, these methods face challenges in accomplishing complex fault diagnosis tasks because the single data source provides very restricted fault features. To address this issue, an attention mechanism‐based multisensor data fusion neural network (MDFNN) for AUV fault diagnosis is proposed in this work. First, a feature extraction layer based on the two‐dimensional (2D) convolutional method with a 1D kernel is introduced to extract features from each sensor data separately, significantly optimizing the model architecture. Second, an efficient channel attention mechanism‐based feature fusion layer is proposed to reassign weights to the features of each sensor data, enabling the model to focus more on crucial features. Finally, the fused features are input to the fully connected layers and softmax layer to realize the fault diagnosis of multisensor data. In the end, the diagnostic performance of the proposed MDFNN is evaluated utilizing real AUV experimental data. The experiment shows that the proposed MDFNN has a very fast convergence speed and 98.37% fault diagnosis accuracy, demonstrating its excellent fault diagnosis performance. The proposed MDFNN provides a generalized and simply structured fault diagnosis framework for the AUV with multiple types of sensor data, providing significant engineering value.