Fault Classification Accuracy Research Articles

An interpretable hybrid framework combining convolution latent vectors with transformer based attention mechanism for rolling element fault detection and classification

Failure of industrial assets can cause financial, operational and safety hazards across different industries. Monitoring their condition is crucial for successful and smooth operations. The colossal volume of sensory data generated and acquired throughout industrial operations supports real-time condition monitoring of these assets. Leveraging digital technologies to analyze acquired data creates an ideal environment for applying advanced data-driven machine learning techniques, such as convolutional neural networks (CNNs) and vision transformer (ViT) to detect faults and classify, enabling accurate prediction and timely maintenance of industrial assets. In this paper, we present a novel hybrid framework based on the local feature extraction ability of CNN with comprehensive understanding of transformer within a global context. The proposed method leverages the complex weight-sharing properties of CNNs and ability of transformers to understand the larger context of spatial relationships in large-scale patterns, making it applicable to datasets of varying sizes. Preprocessing methods such as data augmentation are used to train the model on the Case Western Reserve University (CWRU) dataset in order to increase generalization through computational efficiency. An average fault classification accuracy of 99.62% is accomplished over all three fault classes with an average time-to-fault detection of 38.4 ms. MFPT fault dataset is used to further validate the method with an accuracy of 99.17% for outer race and 99.26% for inner race. Moreover, the proposed framework can be modified to accommodate alternative convolutional models.

A new approach towards more accurate modeling of mechanical defects in power transformer windings

In many cases, power transformers can continue their normal operation despite mechanical defects in their windings. However, estimating the lifespan and assessing the risk of using a transformer with mechanical defects depend on detailed studies of insulation aging and the distribution of transient voltages along the winding, and consequently, the likelihood of partial discharge. Risk assessment studies can only be effectively conducted through accurate modeling of transformers across a wide frequency range. In this study, an attempt has been made to investigate the impact of winding mechanical defects on the transformer equivalent circuits parameters using finite element method precisely modeling of transformer windings, without applying approximation or common simplifications. To compare the effect of each mechanical defect on the resistance, inductance, and capacitance matrices arrays, an index called fault classification accuracy has been used. The investigations have shown that the capacitance matrix arrays undergo the most significant changes for all types of modeled defects in this research. To examine the impact of defects on the transformer frequency response, a π coupled model that directly uses output matrices of finite element modeling has been employed. The sensitivity of various components of frequency response has also been investigated for the used model.

Local fusion generative adversarial network with dual-discriminator and parallel multipath and its application in machinery fault diagnosis with imbalanced data

Abstract Diagnosing faults in critical machinery components is imperative for effective condition monitoring and real-world datasets often suffer from data imbalance. To address this issue, numerous data generation methods have been developed, such as improved local fusion generative adversarial network (ILoFGAN), variational autoencoding GAN (VAEGAN), etc. However, the existing data generation methods primarily concentrate on global and single-scale features and often ignore local or multi-scale features, which leads to the omission of key features or nuances in the generated data. Therefore, a novel approach called the Local Fusion Generative Adversarial Network with Dual-discriminator and Parallel multipath (LoFGAN-DP) is designed to enhance the fault diagnosis performance in the context of imbalanced data. The LoFGAN-DP features a Parallel Multi-Path (PMP) module along with a dual-discriminator scheme, in which the multipath module facilitates feature extraction at various scales through convolution across paths of diverse sizes, and the dual-discriminator scheme can better improve the quality and diversity of the samples generated by the generator. The PMP module and dual-discriminator scheme enhance the proposed method's robustness against variations in input data. After generating data by LoFGAN-DP, a two-dimensional capsule network is further used to achieve the efficient recognition of fault features. To validate the proposed LoFGAN-DP in the machinery fault diagnosis with imbalanced data, the gear dataset and the self-constructed bearing dataset were utilized. Experimental results show that LoFGAN-DP significantly improves Structural Similarity Index (SSIM), Fréchet Inception Distance (FID), and fault classification accuracy compared to several advanced methods.

Transfer learning based cross-process fault diagnosis of industrial robots

In the actual industrial application of robots, the characteristics of robot malfunctions change accordingly as the working environment becomes increasingly diverse and complex. Utilizing the original fault diagnosis models in new working environments correspondingly leads to a decline in the performance and the generalization capability of the model. Moreover, the monitoring data collected in new working processes often has limited or no labels, making the diagnosis models trained with this data unable to identify faults accurately. In this paper, we propose a Domain adaptive Cross-process Fault Diagnosis method (DCFD) to leverage knowledge from existing working processes for diagnosing faults in new working processes. DCFD uses Multi-Kernel Maximum Mean Discrepancy (MK-MMD) to measure the difference between the current working processes and the previous working processes, enhancing the fault diagnosis capability of the robotic system in cross-process scenarios. DCFD achieves an average fault classification accuracy of 98% on 12 types of migration tasks, which demonstrates the effectiveness of DCFD on cross-process fault diagnosis classification tasks in real-time industrial application scenarios.

Improvement of defect detection for steel based on YOLOv7 algorithm

Abstract Working on the conundrum of low efficiency of strip surface defect detection, a surface defect detection algorithm for boards and strips leveraging YOLOV7 is introduced. First, the global attention mechanism is set up to enhance global information interaction and expression capabilities, and improve detection performance; in second place, the C2f component is integrated into the feature pyramid network to the lightweight pattern; ultimately, Focal-EIoU is exerted to replace loss function of the YOLOv7 pattern to solve the divination box and the problem of wrong amplification of length and width, improving the accuracy of defect classification and positioning. The outcomes reveal that the precision of the elevated algorithm has been boosted by about 4.3%, the recall rate has reached 78.3%, and mAP increased by 3.4%. This technique exhibits high accuracy and efficiency in detecting surface defects on the tape.

Electric shock fault identification method based on DWT-AE-BPNN for residual current devices in power distribution systems

The protection dead-zone and threshold setting difficulties of the residual current devices (RCDs) in low-voltage distribution networks may lead to the misidentification of electric shock fault, resulting in severe life-threatening accidents. This paper proposes an electric shock fault identification method based on artificial intelligence for RCDs. Firstly, Mallat discrete wavelet transform (DWT) is applied to efficiently extract non-stationary electric shock feature signals from the total residual current with various noises, preventing weak non-stationary electric shock feature signals from being filtered out. Based on the average and maximum components of the signal mutation, an adaptive threshold can be determined to detect electric shock accurately, avoiding the false activation of RCDs caused by load fluctuations. Subsequently, an autoencoder (AE) is built to mine the non-linear features in which the signal of electric shock on living gradually rises and the signal of electric shock on non-living remains stable. Finally, a back propagation neural network (BPNN) is trained to classify the electric shock types from the non-linear features. The simulation and experiment have been conducted to obtain total residual current data under different conditions, and the electric shock fault real-time identification hardware platforms are developed. The accuracy of electric shock fault detection and classification can reach 100 %, which has advanced its practical applicability.

Fault analysis in clustered microgrids utilizing SVM-CNN and differential protection

The integration of distributed generation, microgrids, and renewable energy sources has significantly enhanced the resilience of modern electrical grids. However, this transition presents challenges in control, stability, safety, and protection due to low fault currents from renewables. This paper addresses these challenges by proposing novel methodologies to enhance fault detection, classification, and localization in microgrids. The literature review highlights a shift towards intelligent learning methods in microgrid protection systems, improving fault response times and identifying electrical faults, including high impedance faults. Nonetheless, existing methods often neglect high impedance fault detection and the integration of differential protection in clustered microgrids. To fill these gaps, this study presents a methodology combining support vector machines and convolutional neural networks for fault detection in microgrids, integrating differential protection for high impedance fault detection. The paper also proposes approaches to optimize protection in clustered microgrid systems. The effectiveness of the methodology is validated using Opal-RT through comparative analyses of signal decomposition techniques, performance and accuracy of support vector machines and convolutional neural networks, K-Fold validation, and sensitivity analysis. Results demonstrate robustness and high performance, achieving up to 100 % accuracy in fault detection and classification.

Hydraulic system fault diagnosis decoupling method based on 2D time-series modeling and self-attention fusion

Hydraulic systems play a pivotal and extensive role in mechanics and energy. However, the performance of intelligent fault diagnosis models for multiple components is often hindered by the complexity, variability, strong hermeticity, intricate structures, and fault concealment in real-world conditions. This study proposes a new approach for hydraulic fault diagnosis that leverages 2D temporal modeling and attention mechanisms for decoupling compound faults and extracting features from multisample rate sensor data. Initially, to address the issue of oversampling in some high-frequency sensors within the dataset, variable frequency data sampling is employed during the data preprocessing stage to resample redundant data. Subsequently, two-dimensional convolution simultaneously captures both the instantaneous and long-term features of the sensor signals for the coupling signals of hydraulic system sensors. Lastly, to address the challenge of feature fusion with multisample rate sensor data, where direct merging of features through maximum or average pooling might dilute crucial information, a feature fusion and decoupling method based on a probabilistic sparse self-attention mechanism is designed, avoiding the issue of long-tail distribution in multisample rate sensor data. Experimental validation showed that the proposed model can effectively utilize samples to achieve accurate fault decoupling and classification for different components, achieving a diagnostic accuracy exceeding 97% and demonstrating robust performance in hydraulic system fault diagnosis under noise conditions.

Fault classification and localization of multi-machine-based ieee benchmark test case power transmission lines using optimizable weighted extreme learning machine

Accurate fault diagnosis in transmission lines is crucial to ensure the reliability and stability of power grids. Conventional approaches often rely on expert knowledge or complex feature extraction methods, which are subjective and time-consuming. Additionally, many existing approaches use separate sub-algorithms for fault classification and localization. These are operating independently and sequentially. This research work proposes an innovative method for fault classification and localization in transmission lines using phasor measurement unit (PMU) data. The proposed method employs a Weighted Extreme Learning Machine (WELM) algorithm, which uses the variable data distribution across different fault classes through a weighted approach. The PMU data is generated by simulation using an IEEE 9-bus test system in the MATLAB simulation environment. The Maximal Overlap Discrete Wavelet Transform-based feature extraction technique is applied to derive input feature data to facilitate fault classification and localization. The WELM classifier is also optimized using the Grey Wolf Optimization (GWO) algorithm. The resulting GWO-optimized WELM (GWO-WELM) model, when trained on PMU data, achieves a remarkable fault classification accuracy of 99.83 % and fault localization accuracy of 95.48 %, respectively. These results demonstrate that the GWO-WELM model outperforms commonly used classifiers. Moreover, the proposed model shows robustness by accurately classifying the noisy data with a signal-to-noise ratio (SNR) of 10 dB (achieving 91.5 % accuracy in classification and 88.1 % accuracy in localization, respectively).

Intelligent fault diagnosis of rolling bearings in strongly noisy environments using graph convolutional networks

SummaryRolling bearings often function under complex and non‐stationary conditions, where significant noise interference complicates fault diagnosis by obscuring fault characteristics. This paper presents an innovative fault diagnosis technique using graph convolutional networks (GCN) to address these challenges. Vibration signals are first transformed into the frequency domain through fast Fourier transform (FFT), creating a detailed graph where nodes and edges encapsulate fault signals. The GCN method then extracts complex node features from this graph, enabling a classifier, comprising a fully connected layer and Softmax function, to accurately identify fault types. Experimental results demonstrate the superior performance of the proposed GCN‐based fault diagnosis method, achieving an accuracy of 99.79%. This significantly surpasses traditional machine learning methods (85.4%), deep learning models (92.3%), and other graph neural network approaches (94.1%). Notably, the method shows exceptional resilience to noise, maintaining high accuracy even with 20% added noise, underscoring its robustness for practical industrial applications. The transformation of vibration signals into the frequency domain using FFT, followed by constructing a detailed graph structure, enables the GCN to effectively capture and represent intricate fault characteristics, thus enhancing accurate fault classification. These findings highlight the method's practical applicability and potential for deployment in advanced industrial settings characterized by high noise levels and complexity.

Fault detection and classification in overhead transmission lines through comprehensive feature extraction using temporal convolution neural network

AbstractFaults in transmission lines cause instability of power system and result in degrading end users sophisticated equipment. Therefore, in case of fault and for the quick restoration of problematic phases, reliable and accurate fault detection and classification techniques are required to categorize the faults in a minimum time. In this work, 500 kV transmission line (Jamshoro‐New Karachi), Sindh, Pakistan has been modeled in MATLAB. The discrete wavelet transform (DWT) has been used to extract features from the transient current signal for different faults in 500 kV transmission line under various parameters such as fault location, fault inception angle, ground resistance and fault resistance and time series data has been obtained for fault classification. Moreover, the temporal convolutional neural network (TCN) is used for fault classification in 500 kV transmission network due to its robust framework. From simulation results, it is found that faults in 500 kV transmission line are classified with 99.9% accuracy. Furthermore, the simulation results of the TCN model compared to bidirectional long short‐term memory (BiLSTM) and Gated Recurrent Unit (GRU) and it has been found that TCN model is capable of classifying faults in 500 kV transmission line with high accuracy due to its ability to handle long receptive field size, less memory requirement and parallel processing due to dilated causal convolutions. Through this work, the meantime to repair of 500 kV transmission line can be reduced.

Digital Twin Modeling for Hydropower System Based on Radio Frequency Identification Data Collection

The safe and steady operation of hydropower generation systems is crucial for electricity output in the grid. However, hydropower stations have complicated interior structures, making defect detection difficult without disassembly inspections. The application of digital modeling to hydropower stations will effectively promote the intelligent transformation of hydropower stations as well as reduce the maintenance costs of the system. This study provides a model of the power generating and transmission system for hydropower plants, with an emphasis on primary equipment and measured data. The model utilizes PSCAD to digitalize state response in hydropower plants with various short-circuit faults. The fault information is identified and learned using the Adaptive Time–Frequency Memory (AD-TFM) deep learning model. It is demonstrated that our proposed method can effectively obtain the fault information through radio frequency identification (RFID). The accuracy of the traditional method is 0.90, while the results for AD-TFM show a fault classification accuracy of 0.92, which is more than enough to identify multiple fault types compared to the existing methods.

BearingFM: Towards a foundation model for bearing fault diagnosis by domain knowledge and contrastive learning

Monitoring bearing failures in production equipment can effectively prevent finished product quality issues and unplanned factory downtime, thereby reducing supply chain uncertainties and risk. Therefore, monitoring bearing failures in production equipment is important for improving supply chain sustainability. Due to the generalization limitations of neural network models, specific models must be trained for specific tasks. However, in real industrial scenarios, there is a severe lack of labeled samples, making it difficult to deploy fault diagnosis models across massive amounts of equipment in workshops. In order to solve the above issue, this paper proposes a cloud-edge-end collaborative semi-supervised learning framework, which provides multi-level computing power and data support for building a foundation model. A data augmentation method based on the bearing fault mechanism is proposed, which effectively preserves the inherent essential characteristics in vibration signals by normalizing frequency and adding noise in specific frequency bands. A novel contrastive learning model is designed, which narrows the distances between positive samples and widens the distances between negative samples in the high-dimensional space through cross comparisons in the time dimension and knowledge dimension, thereby extracting the most essential characteristics from the unlabeled signals. Multiple sets of experiments conducted on four datasets demonstrate that the proposed approach achieves an approximately 98% fault classification accuracy with only 1.2% labeled samples.

An innovative transformer neural network for fault detection and classification for photovoltaic modules

Solar energy from photovoltaic systems (PV) ranks as the third greatest renewable electricity generation resource, expanding quickly through the years as it is free from environmental pollution and has cheap installation costs. Effective performance at high working rates is contingent on the early failure detection of PV modules. This study introduces an innovative deep learning model utilizing a Vision Transformer (ViT) artificial neural network (ANN) for the automatic detection of faults in infrared thermography (IR) images of PV modules. Our approach aims to enhance the accuracy of PV fault detection and classification compared to existing deep learning methods. The proposed framework encompasses three primary stages: (1) image preprocessing, which includes the application of the unsharp mask to sharpen the image’s edges or high-frequency components; (2) data augmentation techniques designed to overcome the problem of unbalanced classes that affect the training process, resulting in learning specific majority classes better than others; and (3) implementing a Vision Transformer deep learning model for its precision in digital image analysis. We evaluated the framework using the public Infrared Solar Modules dataset. The performance was quantitatively assessed using several metrics: accuracy, recall, precision, and F1 score. The dataset is classified into eleven different PV anomalies and another class of no-anomaly PV modules. The results show that our proposed approach has 98.23% accuracy for classifying the dataset into two classes, one for the PV anomaly and the other for the no-anomaly. It also has 96.19% accuracy for classifying eleven PV failures and 95.55% for twelve classes, including the no-anomaly class with the eleven types of anomalies. The experimental results underscore the potential of our model for earlier and more precise detection of PV faults. Furthermore, comparative analysis revealed the superior performance of the proposed approach over other deep learning methods.

A GA based SVM-Bayesian technique and its application in process monitoring for mixed distributed data

Traditional data-driven fault diagnosis methods, such as principal component analysis (PCA) and independent component analysis (ICA), have significantly improved condition monitoring and fault diagnosis of industrial process by analyzing process data. However, their accuracy is limited due to the assumption that the data follows Gaussian or non-Gaussian distribution, respectively. In reality, process data is often a mixture of both Gaussian and non-Gaussian distributions, leading to reduced accuracy of the training model and fault diagnosis results. To address this issue, a genetic algorithm-based SVM-Bayesian (GSB) scheme is proposed in this paper for feature combination and model integration. The scheme first employs ICA and PCA to extract the independent/principal components of the original data, and then constructs the corresponding fault features for SVM training to obtain basic models. The extraction of fault features are optimized using genetic algorithms based on defined accuracy and diversity functions. During online monitoring, the final diagnosis result is obtained by combining the results of the basic models through Bayesian inference. The effectiveness of the proposed scheme is evaluated using a numerical example and a case study of the Tennessee Eastman process. The results show that the accuracy of fault classification is improved by about 5% compared with traditional methods.

An Efficient Double Deep Q Learning Network-Based Soft Faults Detection and Localization in Analog Circuits

To identify whether the circuit under test is fault-free or faulty, fault diagnosis is conducted in an analog circuit. However, in the case of fault detection (FD) techniques, the size of FD is the main challenge in testing, especially for complex analog circuits. Hence to reduce the size of FD, specific test nodes are chosen to perform fault diagnosis from the available accessible test nodes. Specific test nodes are selected based on the faults classification efficiency of fault diagnosis. Therefore, this paper proposes an approach for single and multiple soft fault diagnosis using minimum test nodes in analog electronic circuits. The proposed double deep Q-learning-based hybrid Simulated annealing–Tabu search (DDQN-Hybrid SATS) technique determines minimum test nodes with the utilization of distance metric for fault classification. The DDQN-Hybrid SATS technique is used to identify minimum nodes for testing. In this, the double deep Q learning network (DDQN) ensures better reliability and faster convergence in learning but suffers from catastrophic forgetting issues. To prevent such issues, the DDQN approach is optimized using a hybrid SATS algorithm. The hybrid optimization of simulated annealing (SA) and Tabu search (TS) coalesce the advantages of individual optimization procedures to provide the optimum solution in a fast and effective way. The results for a fourth-order low pass filter are presented (i.e., first CUT) and an eight-bit digital to analog converter (i.e., second CUT). Moreover, the simulation experiment reveals that the proposed DDQN-Hybrid SATS technique achieves greater overall fault classification accuracy of about 96.25% than other compared techniques with minimum computation time.

Circuit Breaker Fault Diagnosis Method Based on Integrated OCSSA-VMD-CNN-BiLSTM Method

Abstract The paper proposes a novel approach to diagnose circuit breaker faults by integrating OCSSA, VMD, CNN, and BiLSTM algorithms. Firstly, OCSSA optimizes the parameters of VMD to reduce noise interference. Then, VMD decomposes the circuit breaker operating signals into characteristic signals. Secondly, CNN captures mechanical fault characteristics from the decomposed signals. Finally, BiLSTM applies CNN’s results for feature classification. Experimental results show that combining improved VMD with the CNN-BiLSTM model achieves a fault classification accuracy rate of 99.5%, surpassing other traditional models such as VMD-CNN-BiLSTM and ELMD in accuracy and robustness.

Deep learning based fault detection of automobile dry clutch system using spectrogram plots

Dry friction clutches are extremely important in the context of power transmission systems. Continuous exposure to extreme heat and loading makes clutch extremely susceptible to various faults. The timely detection and diagnosis of such faults are of utmost importance to prevent any damage to internal components and also helps in avoiding transmission system failures. In this research study, a novel approach that leverages the power of transfer learning (a famous deep learning technique) is proposed to diagnose multiple types of clutch faults including, worn release fingers, fractured pressure plates, deteriorated pressure plates, loss of friction material and distorted tangential strips using spectrogram plots. To train and validate the diagnostic system, vibration readings were taken from a specially designed test rig with the help of piezoelectric accelerometer while the clutch system was operated under different load conditions of 0 (no load), 5 and 10 kg This procedure of data collection was then repeated to acquire the vibration data for all of the fault conditions by replacing the good with fault components individually. These vibration signals were further processed and transformed into spectrogram plots that serves as the input data for the deep learning models considered. Fine-tuning techniques were applied on pretrained networks to maximise the prediction accuracy of the models to effectively determine and diagnose faults in the clutch system. For this study 12 pre-trained networks were chosen namely, Xception, InceptionResNet, DenseNet, AlexNet, VGG16, GoogLeNet, VGG19, ResNet101, ResNet50, InceptionV3, MobileNetV2 and ShuffleNet. To optimize the performance of deep learning models, a systematic adjustment of hyperparameters such as the train-test split ratio, learning rate, optimizer and batch size for each network model was carried out. Through careful experimentation and analysis, significant improvements in fault classification accuracy were achieved thereby enhancing the reliability and effectiveness of the diagnostic system. From the results it was noted that 100% classification accuracy was displayed by AlexNet (for the no load condition and the 10 kg load condition) and GoogLeNet (for 5 kg load condition) with extremely low computation times.

An enhanced semi-supervised learning method with self-supervised and adaptive threshold for fault detection and classification in urban power grids

With the rapid development of urban power grids and the large-scale integration of renewable energy, traditional power grid fault diagnosis techniques struggle to address the complexities of diagnosing faults in intricate power grid systems. Although artificial intelligence technologies offer new solutions for power grid fault diagnosis, the difficulty in acquiring labeled grid data limits the development of AI technologies in this area. In response to these challenges, this study proposes a semi-supervised learning framework with self-supervised and adaptive threshold (SAT-SSL) for fault detection and classification in power grids. Compared to other methods, our method reduces the dependence on labeling data while maintaining high recognition accuracy. First, we utilize frequency domain analysis on power grid data to filter abnormal events, then classify and label these events based on visual features, to creating a power grid dataset. Subsequently, we employ the Yule–Walker algorithm extract features from the power grid data. Then we construct a semi-supervised learning framework, incorporating self-supervised loss and dynamic threshold to enhance information extraction capabilities and adaptability across different scenarios of the model. Finally, the power grid dataset along with two benchmark datasets are used to validate the model’s functionality. The results indicate that our model achieves a low error rate across various scenarios and different amounts of labels. In power grid dataset, When retaining just 5% of the labels, the error rate is only 6.15%, which proves that this method can achieve accurate grid fault detection and classification with a limited amount of labeled data.

Fault diagnosis method via one vs rest evidence classifier considering imprecise feature samples

The key task of fault diagnosis is to establish the nonlinear mapping relationship between fault feature sequences (FFSs) and fault modes (FMs). Therefore, it is usually necessary to detect transit jump points of FFS to separate feature samples with the different trends. These samples and the their fault labels are combined to form training samples for a fault diagnosis model (FDM). However, due to changes in fault status and uncertainty in measurement, some feature samples will imprecisely point to multiple FMs (fault labels), hardly adopted in FDM. In order to solve such imprecision problem, this paper presents an information fusion method via One vs Rest (OvR) evidence classifiers. For each FFS, the multiple OvR evidence classifiers are designed to model the imprecise relationship between the samples and FMs. Then, a two-level fusion framework is proposed to integrate the evidence of all FFSs. By using the intersection operation between One (one FM) and Rest (the other FMs), the imprecision can be significantly reduced. A rotating machinery fault diagnosis experiment is given to illustrate that the proposed method have better fault classification accuracy compared with the traditional FDMs including SVM, BPNN, KNN, FCNN, RF and Bayes.