Fault Classification Performance Research Articles

MDSC-FSPPA-LCFF network for diagnosis of rolling bearing with multipoint fault

Abstract Faults that occur in rolling bearings during operation are complex and variable. While extensive research has been conducted on compound faults involving multiple components, studies on multiple faults in single component are relatively scarce. However, the occurrence of multiple faults in single component is a common phenomenon. To address the issues of difficulty in feature extraction, numerous network parameters, and slow computational speed, a multi-scale dynamic snake convolution with fast spatial pyramid pooling attention (MDSC-FSPPA) and lightweight comprehensive feature fusion (LCFF) network is proposed for multipoint fault diagnosis of rolling bearings. Firstly, multi-scale shallow feature extraction (MSFE) module is applied to extract the features from the original signals. Then, dynamic snake convolution (DSC) with FSPPA module is used to refine these features deeply. Subsequently, LCFF module is employed to reduce network parameters while still fully extracting fault features. Additionally, fault identification is obtained through the softmax function. Finally, the t-distributed stochastic neighbor embedding (t-SNE) method is utilized to visually demonstrate the fault classification performance of the proposed method. The experimental evaluation conducted on bearing datasets indicates that the proposed network exhibits excellent performance of multipoint fault diagnosis in rolling bearings.

Sensor Fault Detection and Classification Using Multi-Step-Ahead Prediction with an Long Short-Term Memoery (LSTM) Autoencoder

The Internet of Things (IoT) is witnessing a surge in sensor-equipped devices. The data generated by these IoT devices serve as a critical foundation for informed decision-making, real-time insights, and innovative solutions across various applications in everyday life. However, data reliability is often compromised due to the vulnerability of sensors to faults arising from harsh operational conditions that can adversely affect the subsequent operations that depend on the collected data. Hence, the identification of anomalies within sensor-derived data holds significant importance in the IoT context. This article proposes a sensor fault detection method using a Long Short-Term Memory autoencoder (LSTM-AE). The AE, trained on normal sensor data, predicts a 20-step window, generating three statistical features via SHapley Additive exPlanations from the estimated steps. These features aid in determining potential faults in the predicted steps using a machine learning classifier. A secondary classifier identifies the type of fault in the sensor signal. Experimentation on two sensor datasets showcases the method’s functionality, achieving fault detection accuracies of approximately 93% and 97%. It is possible to attain a perfect fault classification performance by slightly modifying the feature calculation approach. In a univariate prediction scenario, our proposed approach demonstrates good fault detection and classification performance.

A novel domain adaptive method for gearbox fault diagnosis using maximum multiple-classifier discrepancy network

Abstract Domain adaptive gearbox fault diagnosis methods have made impressive achievements for the past several years. However, most of the traditional domain adaptive methods have significant limitations under fluctuating operating conditions. The acquired acceleration signals will result in different signal vibration spectra and peak vibration amplitudes due to the different working conditions between the source and target domains. There is an obvious discrepancy between the distribution of fault samples in the source domain and the target domain, which in turn makes it difficult to classify the target domain samples with fuzzy fault category boundaries. Therefore, how to measure the discrepancy between two distributions has been an important research direction in machine learning. A good metric helps to discover better features and build better models. In this paper, a novel domain adaptive method for gearbox fault diagnosis using maximum multiple-classifier discrepancy network (MMCDN) is proposed. The sparse stack autoencoder is used by the MMCDN as a feature extractor for fault feature extraction, and a kind of composite distance is adopted for domain discrepancy measurement of source and target domain features for domain alignment. Then the extracted features are input into a three-classifier of the model for adversarial training. The trained model classifiers have high performance in fault classification. The combination of domain adaptation and multi-classifier discrepancy output can effectively solve the impact of working condition changes and the misclassification problem for fuzzy samples with class boundaries. Experimental validation with two planetary gearbox datasets shows that the MMCDN has more favorable diagnostic accuracy and performance than other methods.

Ensemble-based deep learning model for welding defect detection and classification

This study introduces an ensemble-based deep learning approach for monitoring and detecting submerged arc weld defects in weld beads and adjacent zones during Non-Destructive Testing (NDT). The methodology utilizes an open-source image dataset to categorize submerged arc weld conditions into three defect types: cracks, lack of penetration and porosity, and ideal welds (no defects). To enhance image analysis, we employ preprocessing and feature extraction in both spatial and frequency domains using segmentation techniques like grey level difference method (GLDM), grey level co-occurrence matrix (GLCM), discrete wavelet transform (DWT), Fast Fourier Transform (FFT), and texture analysis. Following image preparation, the preprocessed data undergoes training and testing in our proposed ensemble-based deep learning model. The model's performance is thoroughly assessed using key metrics such as precision, recall, F1 score, receiver operating characteristics (ROC) curve, and a confusion matrix, resulting in an impressive accuracy rate of 93.12% for detecting and classifying weld faults. In comparing our model to state-of-the-art techniques in existing literature, our ensemble-based deep learning approach consistently demonstrates a high fault detection and classification performance in the face of a simplistic MLP-based core architecture, ratifying the robustness of our feature ensemble approach. The study concludes that this model has significant potential for integration into current inspection systems, offering an efficient and robust solution for real-time condition monitoring of welds along weldment joints.

Implementing generative adversarial networks for increasing performance of transmission fault classification

<p>An electrical power system is a network that facilitates the sourcing, transfer, and distribution of electrical energy. In the traditional power system, there are eleven types of faults that can occur in the system. This paper focuses on the classification of these faults over a stretch of 100 kilometres. The dataset used is synthetic and generated from a simulated model using MATLAB/Simulink software. Data augmentation is carried out during training to improve the accuracy of the classification. An indirect training approach through generative adversarial network (GAN) is used to classify these overhead transmission line faults. The random forest (RF) classification is used as the base learning model on the original dataset and it achieves accuracy of 84%. However, the base learner RF when used on GAN model generated augmented faulty data, it performs exceptionally well achieving 99% accuracy. One of the recent state-of-art methods is compared with this approach.</p>

A multilayer integrative approach for diagnosis, classification and severity detection of electrical faults in photovoltaic arrays

Early detection of faults in photovoltaic (PV) systems is crucial for improving their lifespan and reliability. Conventional protection devices may not be able to detect electrical faults under critical conditions. This usually occurs due to (i) the current-limiting nature and non-linear output characteristics of PV arrays, (ii) the changing environmental conditions (such as irradiance and temperature variations), and (iii) the influence of the maximum power point tracker (MPPT), particularly in the presence of critical fault impedance values or mismatch levels. Therefore, modern data-driven methods are required for fault detection and classification in PV arrays. However, careful investigation in previous studies reveals the existing gaps and limitations, such as poor accuracy, and incomprehensive models which have not considered various electrical faults, low mismatch levels, and critical fault impedance values. To this end, in this study, a comprehensive multilayer model is proposed which consists of six layers for diagnosis, classification and severity identification of electrical faults in PV arrays. Each layer adopts a weighted ensemble learning (WEL) algorithm consisting of three classifiers namely Support Vector Machine (SVM), Naïve Bayes (NB) and Logistic Regression (LR). Besides, a genetic algorithm (GA) is employed in each layer to find the optimal weights for each classifier (i.e., SVM, NB and LR). The initial dataset is acquired through an investigation into the PV array current–voltage (I-V) characteristic curve and extraction of several features under various faulty and normal conditions. To reduce the dataset dimensionality thus computationally simplifying the training process, the sequential floating forward selection (SFFS) algorithm is utilized in each layer as a powerful feature selection technique. The results show a highly accurate performance in fault diagnosis, classification, and severity assessment, with an average simulation and experimental accuracies of 98.9% and 98.37%, respectively.

A Multi-Input Convolutional Neural Network Model for Electric Motor Mechanical Fault Classification Using Multiple Image Transformation and Merging Methods

This paper addresses the critical issue of fault detection and prediction in electric motor machinery, a prevalent challenge in industrial applications. Faults in these machines, stemming from mechanical or electrical issues, often lead to performance degradation or malfunctions, manifesting as abnormal signals in vibrations or currents. Our research focuses on enhancing the accuracy of fault classification in electric motor facilities, employing innovative image transformation methods—recurrence plots (RPs), the Gramian angular summation field (GASF), and the Gramian angular difference field (GADF)—in conjunction with a multi-input convolutional neural network (CNN) model. We conducted comprehensive experiments using datasets encompassing four types of machinery components: bearings, belts, shafts, and rotors. The results reveal that our multi-input CNN model exhibits exceptional performance in fault classification across all machinery types, significantly outperforming traditional single-input models. This study not only demonstrates the efficacy of advanced image transformation techniques in fault detection but also underscores the potential of multi-input CNN models in industrial fault diagnosis, paving the way for more reliable and efficient monitoring of electric motor machinery.

Fault Classification in Reciprocating Compressors: A Comparison of Machine Learning and Deep Learning Approaches

This study compares methodologies for fault classification in reciprocating compressors, focusing on traditional Machine Learning (ML) with classical feature extraction processes and one-dimensional Convolutional Neural Networks (1D-CNN) in Deep Learning (DL). Both techniques demonstrated viability by employing a dataset of compressor vibration signals encompassing ten fault classes. While ML achieved a classification accuracy of 86%, DL reached 90.709%, highlighting its superior learning and generalization abilities, although with longer training times. These findings suggest that, despite ML being effective when relevant prior knowledge is available, DL, particularly with 1D-CNN, offers enhanced fault classification performance for this study case at the expense of additional processing resources.

Confidentiality-preserving machine learning algorithms for soft-failure detection in optical communication networks

Automated fault management is at the forefront of next-generation optical communication networks. The increase in complexity of modern networks has triggered the need for programmable and software-driven architectures to support the operation of agile and self-managed systems. In these scenarios, the European Telecommunications Standards Institute zero-touch network and service management approach is imperative. The need for machine learning algorithms to process the large volume of telemetry data brings safety concerns as distributed cloud-computing solutions become the preferred approach for deploying reliable communication network automation. This paper’s contribution is twofold. First, we propose a simple yet effective method to guarantee the confidentiality of the telemetry data based on feature scrambling. The method allows the operation of third-party computational services without direct access to the full content of the collected data. Additionally, the effectiveness of four unsupervised machine learning algorithms for soft-failure detection is evaluated when applied to the scrambled telemetry data. The methods are based on factor analysis, principal component analysis, nonlinear principal component analysis, and singular value decomposition. Most dimensionality reduction algorithms have the common property that they can maintain similar levels of fault classification performance while hiding the data structure from unauthorized access. Evaluations of the proposed algorithms demonstrate this capability.

Performance evaluation of three signal decomposition methods for bearing fault detection and classification

In the present study, the performance evaluation of the signal decomposition methods; variational mode decomposition, empirical mode decomposition, and ensemble empirical mode decomposition, for the ball bearing fault detection and classification for the experimentally recorded vibration signals has been done. This work proposed a novel hybrid sensitive mode selection method combining three statistical measures (energy-based index, fault correlation-based index, and Hausdorff distance-based index) and investigating the effect of the selected sensitive mode extracted by the decomposition methods for the bearing defect frequency detection. The vibration data have been acquired for the healthy and seeded faults of different sizes for the inner and outer raceway defects. The complete features dataset comprises five time-domain, four spectral-domain, and two non-linear statistical features. The k-Nearest Neighbor, Support Vector Machine, and Naive Bayes classifiers are used for fault classification and predict the results with four performance metrics: accuracy, sensitivity, precision, and F-score. Firstly, the results of signal decomposition employing hybrid sensitive mode functions and statistical analysis of condition indicators (RMS, kurtosis and crest factor) revealed that the VMD outperforms the other two techniques. Secondly, the fault classification results predicted that the k-Nearest Neighbor classifier outperforms the other two classifiers. This proposed novel sensitive mode selection method significantly improves the bearing fault classification performance metrics with the features extracted from the selective mode functions with all three decomposition methods.

A Predictive Maintenance System Design and Implementation for Intelligent Manufacturing

The importance of predictive maintenance (PdM) programs has been recognized across many industries. Seamless integration of the PdM program into today’s manufacturing execution systems requires a scalable and generic system design and a set of key performance indicators (KPIs) to make condition monitoring and PdM activities more effective. In this study, a new PdM system and its implementation are presented. KPIs and metrics are proposed and implemented during the design to enhance the system and the PdM performance monitoring needs. The proposed system has been tested in two independent use cases (autonomous transfer vehicle and electric motor) for condition monitoring applications to detect incipient equipment faults or operational anomalies. Machine learning-based data augmentation tools and models are introduced and automated with state-of-the-art AutoML and workflow automation technologies to increase the system’s data collection and data-driven fault classification performance.

A Hard Voting Policy-Driven Deep Learning Architectural Ensemble Strategy for Industrial Products Defect Recognition and Classification.

Manual or traditional industrial product inspection and defect-recognition models have some limitations, including process complexity, time-consuming, error-prone, and expensiveness. These issues negatively impact the quality control processes. Therefore, an efficient, rapid, and intelligent model is required to improve industrial products’ production fault recognition and classification for optimal visual inspections and quality control. However, intelligent models obtained with a tradeoff of high accuracy for high latency are tedious for real-time implementation and inferencing. This work proposes an ensemble deep-leaning architectural framework based on a deep learning model architectural voting policy to compute and learn the hierarchical and high-level features in industrial artefacts. The voting policy is formulated with respect to three crucial viable model characteristics: model optimality, efficiency, and performance accuracy. In the study, three publicly available industrial produce datasets were used for the proposed model’s various experiments and validation process, with remarkable results recorded, demonstrating a significant increase in fault recognition and classification performance in industrial products. In the study, three publicly available industrial produce datasets were used for the proposed model’s various experiments and validation process, with remarkable results recorded, demonstrating a significant increase in fault recognition and classification performance in industrial products.

Uncertainty awareness in transmission line fault analysis: A deep learning based approach

With the expansion of the modern power system, it is of increasing significance to analyze the faults in the transmission lines. As the transmission line is the most exposed element of a power system, it is prone to different types of environmental as well as measurement uncertainties. This uncertainties influence the sampled signals and negatively affects the fault detection and classification performance. Therefore, an unsupervised deep learning framework named deep belief network is presented in this paper for fault detection and classification of power transmission lines. The proposed framework learns the beneficial feature information from the uncertainty affected signals with a unique two stage learning strategy. This strategy enables the proposed framework to extract lower level fault-oriented information which may remain unobserved for other alternative approaches. The efficacy of the proposed framework has been examined on the IEEE-39 bus benchmark topology. The in-depth accuracy assessment with different accuracy metrics along with exclusive case studies such as the influence of noise, measurement error as well as line and source parameter variations will be conducted in this paper to justify the real-world applicability of the proposed framework. Furthermore, the relative performance assessment with the cutting-edge rival techniques is also presented in this paper to verify if the proposed framework attains a state-of-the-art classification performance or not.

Inter-Relational Mahalanobis SAE with semi-supervised strategy for fault classification in chemical processes

Since industrial process data often presents strong correlations, high complexity, and nonlinear patterns, a proficient deep learning model is required for the fault classification task. Recent researches have shown that deep learning models like stacked autoencoder (SAE) are able to learn deep abstract features from complex process data. Nevertheless, a traditional SAE cannot extract the informative fault-relevant features and data distribution features from industrial process data, which are necessary for effective fault classification in industrial processes. Thus, this study proposes a semi-supervised Inter-Relational Mahalanobis SAE (IRM-SAE) model to learn inter-relational distribution and fault-relevant dynamic features of process data for fault classification. First, the Inter-Relational Mahalanobis loss is introduced into the original objective function of SAE to learn meaningful inter-relational distribution features within the data. Then, an active time frame technique is developed to preprocess the input data to capture the dynamic features of the data. Furthermore, to fully utilize both labeled and unlabeled data in industrial processes, the semi-supervised strategy is introduced to learn fault-related features for better fault classification. To validate the performance of the proposed model, it is applied on the Tennessee–Eastman process and a real-world industrial hydrocracking process. The experimental results show that the proposed model has higher fault classification performance compared to other deep learning models.

High-Performance Fault Classification Based on Feature Importance Ranking-XgBoost Approach with Feature Selection of Redundant Sensor Data

Background: Through the analysis of the relevant data of industrial equipment, faults diagnosis is helpful for system maintenance and reducing economic losses. Objective: The aim is to reduce the influence of irrelevant features and efficiently train the FIR-XgBoost model. Methods: An Extreme Gradient Boosting (XgBoost) approach based on feature importance ranking (FIR) is proposed in this article for fault classification of high- dimensional complex industrial systems. Gini index is applied to rank the importance of the features, and feature selection is implemented based on their position in the ranking. Results: The dataset from the PHM 2021 data challenge, which is related to the process of fuse thermal imaging, is used. The classification accuracy of FIR-XgBoost reaches 99.63%, outperforming other existing algorithms. A case study is presented to show that excellent fault classification can be achieved through ensemble learning and feature selection. Conclusion: Data-driven machine learning methods are proposed for solving high-dimensional fault classification problems on the dataset of the PHM2021 Data Challenge. An FIR-XgBoost method is proposed, the core of which is to retain important features and to reduce redundancy of sensor data. Consequently, feature selection based on FIR has better interpretability than other algorithms. Furthermore, the FIR- XgBoost algorithm retaining the 50 most important features achieves the best fault classification performance among the compared algorithms and can be implemented in specific industrial processes.

상호 다중 모델을 이용한 멀티로터 프로펠러 고장 감지

Many studies are being promoted due to the miniaturization of equipment installed in the UAV(Unmanned Aerial Vehicle). Among them, the quadrotor of MAV(Multirotor Aerial vehicle) is the most popular UAV and has been developed in various fields. Practical utilization of UAV requires stable flight or emergency action in fault of UAV parts, but a quadrotor with one motor damaged cannot perform stable flight due to unbalanced force applied to the quadrotor frame. For this reason the development of hexa- or octorotor with more motors is underway and the need to develop FDI(Fault Diagnosis and Isolation) algorithms to use with MAV is increasing. In this paper we designed a simulation based on MAV dynamic and FDI used to classify fault that occurred in the simulation. For fault diagnosis the model based FDI algorithm IMM(Interacting Multiple Model) was applied because the fault affects the system dynamic and output. IMM uses multiple filter models in simultaneously and then compare to the filters estimation and target system output to give weight to suitable filter model. IMM showed accurate and fast fault perceive and classification performance when a fault occurred in the MAV simulation.

Multi-component fault classification of a wind turbine gearbox using integrated condition monitoring and hybrid ensemble method approach

Gearbox has a high failure rate and downtime per failure in a wind turbine. Condition monitoring (CM) helps in diagnosis and prognosis of the faults. CM uses sensors to acquire data from different components and these signals are processed to extract fault features. The extracted features are further used to train different data driven models to identify and quantify the faults. A single decision-making algorithm does not perform well in different scenarios. To overcome this limitation, ensemble methods are used to diagnose the faults. Bagging and boosting are the widely used ensemble methods. However, boosting algorithms performs well on a noise-free data compared to bagging and in noisy settings, bagging methods are much robust than boosting. For this reason, a hybrid ensemble method that combines boosting, bagging, and stacking techniques to identify faults in a gearbox is proposed. Random forest (RF), decision tress (DT), KNN, AdaBoost, and gradient booster were used to compare the fault classification performance of the developed method. Experiments were conducted on the gearbox considering multi-component faults and integrated condition monitoring scheme. Vibration and acoustic signals acquired for different fault cases were processed using discrete wavelet transform (DWT) and the feature extraction was performed to obtain Kurtosis and Shannon entropy. These features were used to classify the faults using aforementioned algorithms. Results indicate that the developed method has the highest fault classification accuracy of 92%. The proposed method can be further used to diagnose similar machine faults in the industries.

Research on Improved Deep Convolutional Generative Adversarial Networks for Insufficient Samples of Gas Turbine Rotor System Fault Diagnosis

In gas turbine rotor systems, an intelligent data-driven fault diagnosis method is an important means to monitor the health status of the gas turbine, and it is necessary to obtain sufficient fault data to train the intelligent diagnosis model. In the actual operation of a gas turbine, the collected gas turbine fault data are limited, and the small and imbalanced fault samples seriously affect the accuracy of the fault diagnosis method. Focusing on the imbalance of gas turbine fault data, an Improved Deep Convolutional Generative Adversarial Network (Improved DCGAN) suitable for gas turbine signals is proposed here, and a structural optimization of the generator and a gradient penalty improvement in the loss function are introduced to generate effective fault data and improve the classification accuracy. The experimental results of the gas turbine test bench demonstrate that the proposed method can generate effective fault samples as a supplementary set of fault samples to balance the dataset, effectively improve the fault classification and diagnosis performance of gas turbine rotors in the case of small samples, and provide an effective method for gas turbine fault diagnosis.

A novel multivariate statistical process monitoring algorithm: Orthonormal subspace analysis

Partial least squares (PLS) and canonical correlation analysis (CCA) are two most popular key performance indicators (KPI) monitoring algorithms, which have shortcomings in dealing with the KPI-related information leakage problem, the model identification problem, the component selection problem, and the process hypothesis problem. To overcome these shortcomings, this study proposes a new multivariate statistics-based process monitoring (MSPM) method called the orthonormal subspace analysis (OSA) method. OSA divides process data and KPI data into three orthonormal subspaces through an analytic solution. Hence, OSA not only can detect a fault but also can judge whether the fault is KPI-related or KPI-unrelated. OSA is always effective irrespective of the availability of the KPI variables in the online monitoring stage. In OSA, the cumulative percent variance (CPV) method is adopted for principal component selection. In brief, OSA is a more effective method than PLS and CCA, and it achieves superior performance in fault detection and classification.

Fault Diagnosis of Rolling Bearings Using Dual-Tree Complex Wavelet Packet Transform and Time-Shifted Multiscale Range Entropy

Most existing fault diagnosis methods for rolling bearings are single-stage; these methods can only judge the fault type but cannot detect the existence of a fault. Moreover, the uncertainty in pattern recognition may lead to misclassification of healthy bearings as faulty ones. This paper proposes a multistage fault detection scheme for rolling bearings. In the first stage, the sensitivity of the range entropy to bearing failure is used to define a threshold, based on which the health status of the bearing is judged. If the unknown bearing is judged to be faulty, the next stage is implemented. In the second stage, a fault feature extraction method based on dual-tree complex wavelet packet transform (DTCWPT), time-shifted multiscale range entropy (TSMRE), and t-distributed stochastic neighbor embedding (t-SNE) is proposed, and a random forest (RF) discriminator is used for fault classification. To achieve the desired performance of fault classification, a new coarsening approach for complexity measurement called TSMRE is developed on the basis of the range entropy (RE). First, the RE value of each time-shifted coarse-grained time series is calculated, and the TSMRE is obtained by averaging the entropy values. The TSMRE improves the coarse-graining processing of the MRE and enhances the stability and reliability of the algorithm. In addition, it can obtain more information from short time series using the time-shifted coarse-grained technology. Therefore, it is less dependent on the length of the original time series. Two sets of rolling bearing data are used for this experiment. The fault recognition rate of each category of samples is 100%. Therefore, the proposed multistage fault diagnosis method can pre-screen healthy bearings and accurately identify the failure types of faulty bearings.