Recursive Spectral Meta-Learner for Online Combining Different Fault Classifiers

The problem of fault classification in industry has been studied extensively. Most classification algorithms are modeled on the premise of data balance. However, the difficulty of collecting industrial data in different modes is quite different. This inevitably leads to data imbalance, which will adversely affect the fault classification performance. This article proposes a novel data augmentation classifier (DAC) for imbalanced fault classification. Data augmentation based on generative adversarial networks (GANs) is an effective way to solve the problem of unbalanced classification. However, the randomness of the GAN generation process restricts the effect of data enhancement. DAC proposes a data selection strategy based on data filtering and data purification in model training to solve this problem. In addition, DAC combines supervised learning and data generation processes to obtain an end-to-end model. Meanwhile, multigenerator structure of DAC (MDAC) is proposed to solve the problem of incomplete learning of a single generator when data imbalances get complicated. The proposed DAC and MDAC are applied in two fault classification cases of the Tennessee Eastman (TE) benchmark process, results of which show superiority of DAC and MDAC compared to existing methods. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —Data imbalances are common in fault classification and affect the effectiveness of modeling in industry. As a generative model, generative adversarial networks (GANs) provide new ideas for small-class data augmentation. However, the instability of its training process and the randomness of data generation affect the results of data augmentation. In this article, the GAN generation process is analyzed in detail. The results of the visualization indicate that no data generation was perfect at any one time. Based on the rules of GAN data generation, we propose a data selection strategy during training. High-quality data are selected for data augmentation through data filtering and data purification. Apart from this, we combine the training process of GAN and classification model for imbalanced data to reduce modeling time. Through industrial examples, we have evaluated the effectiveness of this method.

Fuzzy decision fusion system for fault classification with analytic hierarchy process approach

Discriminative common vector in sufficient data Case: A fault detection and classification application on photovoltaic arrays

Switching LDS-based approach for process fault detection and classification

A modified fuzzy min–max neural network with rule extraction and its application to fault detection and classification

This paper presents a modular yet integrated approach to the problem of fast fault detection and classification. Although the specific application example studied here is a power system, the method would be applicable to arbitrary dynamic systems. The approach is quite flexible in the sense that it can be model-based or model-free. In the model-free case, we emphasize the use of concepts from signal processing and wavelet theory to create fast and sensitive fault indicators. If a model is available then conventionally generated residuals can serve as fault indicators. The indicators can then be analyzed by standard statistical hypothesis testing or by artificial neural networks to create intelligent decision rules. After a detection, the fault indicator is processed by a Kohonen network to classify the fault. The approach described here is expected to be of wide applicability. Results of computer experiments with simulated faulty transmission lines are included.

Power system fault type classification and location prediction is critical in assessing the reliability of the power system, and later restoring it to a stable operating point followed by a fault. State of the art methods include sequence component, impedance measurement from the origin of the fault, and traveling wave based methods for fault type detection and classification problem. It is important to identify and classify the fault as quickly as possible for restoring power system stability to normal operation. Machine Learning and Deep Learning methods allow the analysis of large data of fault voltages and currents by using fast and efficient algorithms. These methods require large amount of data, however, with the recent advances in the field of power system, data acquisition using smart meters and Phasor Measurement Units (PMU), huge amount of system-wide data can be made available to analyze the problem of fault type classification and location prediction. This paper presents a comparative study of Stochastic Gradient Descent (SGD) based Deep Neural Network (DNN) and Machine Learning (ML) applied to power system fault type and location prediction problem. DNN architecture uses 10 hidden layers, each layer having 60 units with hyperbolic tangent as activation function, and a combination of Support Vector Machine (SVM) and Principal Component Analysis (PCA) method are considered. Comparative results in terms of time taken to run the algorithms, and accuracy of the results obtained are presented for a 3 machine 9 bus system. Results indicate that SVM method is an optimal choice for fault classification with high accuracy for location prediction and low computational requirements compared with DNN.

Industrial control systems rely on alarm systems to alert operators to abnormal conditions. These systems frequently generate alarm floods, periods of intense alarm activities that overwhelm operators and compromise situational awareness. The root causes of alarm floods range from operational faults arising from equipment failures to sophisticated cyberattacks deliberately designed to manipulate process controls. Despite their appearance as periods of intense alarm activities, alarm floods resulting from operational faults and cyberattacks are fundamentally different. Recognition of this distinction is essential, as each category requires its own diagnostic approach and response protocols. This thesis addresses the dual challenge of classifying alarm floods and identifying their originating fault types. It also detects anomalous patterns indicative of previously unseen cyber intrusions. This research presents two complementary deep learning frameworks for alarm flood analysis. First, a convolutional neural network (CNN) architecture is developed for online alarm flood classification that transforms temporal alarm sequences into spatial-temporal heatmap representations. This visual pattern recognition system captures alarm activation patterns across process variables and time, enabling real-time fault classification. The framework incorporates multi-channel input encoding to distinguish valid alarm data from padded regions, progressive feature extraction through hierarchical convolutional layers, and demonstrates early classification capability using partial alarm information as floods evolve. Second, a class-conditioned autoencoder (C2AE) framework is introduced for open-set recognition, enabling the system to distinguish between known operational faults and previously unseen cyberattack patterns. C2AE employs a maximum separation training strategy that explicitly maximizes the reconstruction quality gap between matched and mismatched class conditions. This approach enables robust detection of cyberattacks as out-of-distribution samples through reconstruction-based anomaly scoring, while simultaneously classifying known faults. The framework addresses the critical limitation of closed-set classifiers that force unknown cyberattack patterns into known fault categories. Both methodologies are validated on the Tennessee Eastman Process benchmark, a widely adopted testbed for process control research. The CNN-based classifier achieves high accuracy in distinguishing between five operational fault scenarios, with classification performance improving as alarm floods evolve over time. The C2AE framework demonstrates effective open-set recognition by successfully identifying four types of cyber attacks as unknown patterns while maintaining accurate classification of known operational faults. Comprehensive evaluation includes closed-set fault classification performance, open-set cyber attack detection capabilities, and analysis of reconstruction error patterns that enable discrimination between legitimate faults and malicious intrusions. The proposed frameworks provide operators with enhanced situational awareness and enable appropriate emergency response strategies in safety-critical industrial environments facing both operational disturbances and evolving cyber security threats.

In this paper, we address the problem of fault classification in PhotoVoltaic (PV) arrays using a semi-supervised graph signal processing approach. Traditional fault detection and classification methods require large amounts of labeled data for training. In utility scale solar arrays, obtaining labeled data for different fault classes is resource intensive. We propose a graph based classification technique that relies on a limited amount of labeled data. We compare our results with the well known supervised machine learning classifiers such as the K-nearest neighbour classifier, random forest classifier, support vector machines, and artificial neural networks. We also show that the graph-based classifiers require lower training computational cost compared to the standard supervised machine learning algorithms. The proposed method also achieves good classification performance with unseen data. We validate our method on a real-time dataset and show significant improvements over existing approaches.

Using warping information for batch process monitoring and fault classification

For the problem of classification and identification of defects in polyethylene (PE) gas pipelines, this paper firstly performs preliminary screening of the acquired images and acquisition efficiency of defective image acquisition was improved. Images of defective PE gas pipelines were pre-processed. Then, edge detection of the defective images was performed using the improved Sobel algorithm and an adaptive threshold segmentation method was applied to segment the defects in the pipeline images. Finally, the defect images were morphologically processed to obtain binary images. The obtained binary images were applied with VGG16 to complete the training of the defect classifier. The experimental findings show that in the TensorFlow API environment, the test set’s highest accuracy reached 97%, which can achieve the identification of defect types of underground PE gas transmission pipelines.

With modern health monitoring systems on-board, helicopters are still experiencing undetected faults in super critical mechanical components. Two fatal crashes of the Airbus Super Puma helicopter were caused by faults in a planetary gear-bearing component. The accident investigations showed that the fault (i.e. the bore crack) initiated from fatigue spalls on the planet bearing outer raceway, which is integral with the bore of the planet gear. A recent study proposed a method based on planet-gear synchronous signal averaging to detect a planet gear bore crack. Applying deep-learning networks to gear fault diagnosis has becomes a popular topic in machine health monitoring, with the majority of studies focused on fault classification problems. In this paper, we propose a method that relies mostly on the data without detailed knowledge of the system physics. We employ a deep machine learning method, i.e. the Long Short-Term Memory based Recurrent Neural Network (LSTM-RNN), as a signal processing tool to analyse vibration data generated in a gear rig test with faulty gears. We focus on detecting anomalies or changes induced by the gear faults in the LSTM prediction error signal, or LSTM-Residual. We compare the fault detectability of LSTM-Residual with that of some of the commonly used residual signals in gear fault diagnostics. We also provide a detailed study on how to apply the LSTM to monitoring gear tooth crack growth at various stages of the fault development, and what the potential pitfalls are in applying the LSTM network—especially the problem of false detection.

Principal component analysis (PCA) is a linear data analysis technique widely used for fault detection and isolation, data modeling, and noise filtration. PCA may be combined with statistical hypothesis testing methods, such as the generalized likelihood ratio (GLR) technique in order to detect faults. GLR functions by using the concept of maximum likelihood estimation (MLE) in order to maximize the detection rate for a fixed false alarm rate. The benchmark Tennessee Eastman Process (TEP) is used to examine the performance of the different techniques, and the results show that for processes that experience both shifts in the mean and/or variance, the best performance is achieved by independently monitoring the mean and variance using two separate GLR charts, rather than simultaneously monitoring them using a single chart. Moreover, single-valued data can be aggregated into interval form in order to provide a more robust model with improved fault detection performance using PCA and GLR. The TEP example is used once more in order to demonstrate the effectiveness of using of interval-valued data over single-valued data.

An explainable artificial intelligence based approach for interpretation of fault classification results from deep neural networks

A single Bayesian network classifier for monitoring with unknown classes