Fault Type Classification Research Articles

Classification of Faults in Power System Transmission Lines Using Deep Learning Methods with Real, Synthetic, and Public Datasets

Every part of society relies on energy systems due to the growing population and the constant demand for energy. Because of the high energy demands of transportation, industry, and daily life, energy systems are crucial to every part of society. Electrical transmission lines are a crucial component of the electrical power system. Therefore, in order to determine the power system’s protection plan and increase its reliability, it is critical to foresee and classify fault types. With this motivation, the main goal of this paper is to design a deep network model to classify faults in transmission lines based on real, generated, and publicly available datasets. A deep learning architecture that was based on a one-dimensional convolutional neural network (CNN) was utilized in this study. Accuracy, specificity, recall, precision, F1 score, ROC curves, and AUC were employed as performance criteria for the suggested model. Not only synthetic but also real data were used in this study. It has been seen that the created model can be used successfully for both real data and synthetic data. In order to measure the robustness of the network, it was tested with three different datasets consisting of real, generated, and publicly available datasets. In the paper, 1D CNN, one of the machine learning methods, was used on three different power systems, and it was observed that the machine learning method was successful in all three power systems.

Bearing fault detection in adjustable speed drives via self-organized operational neural networks

AbstractAdjustable speed drives (ASDs) are widely used in industry for controlling electric motors in applications such as rolling mills, compressors, fans, and pumps. Condition monitoring of ASD-fed induction machines is very critical for preventing failures. Motor current signature analysis offers a non-invasive approach to assess motor condition. Application of conventional convolutional neural networks provides good results in detecting and classifying fault types for utility line-fed motors, but the accuracy drops considerably in the case of ASD-fed motors. This work introduces the use of self-organized operational neural networks to enhance the accuracy of detecting and classifying bearing faults in ASD-fed induction machines. Our approach leverages the nonlinear neurons and self-organizing capabilities of self-organized operational neural networks to better handle the non-stationary nature of ASD operations, providing more reliable fault detection and classification with minimal preprocessing and low complexity, using raw motor current data.

Transformer-based deep learning networks for fault detection, classification, and location prediction in transmission lines

ABSTRACT Fault detection, classification, and location prediction are crucial for maintaining the stability and reliability of modern power systems, reducing economic losses, and enhancing system protection sensitivity. This paper presents a novel Hierarchical Deep Learning Approach (HDLA) for accurate and efficient fault diagnosis in transmission lines. HDLA leverages two-stage transformer-based classification and regression models to perform Fault Detection (FD), Fault Type Classification (FTC), and Fault Location Prediction (FLP) directly from synchronized raw three-phase current and voltage samples. By bypassing the need for feature extraction, HDLA significantly reduces computational complexity while achieving superior performance compared to existing deep learning methods. The efficacy of HDLA is validated on a comprehensive dataset encompassing various fault scenarios with diverse types, locations, resistances, inception angles, and noise levels. The results demonstrate significant improvements in accuracy, recall, precision, and F1-score metrics for classification, and Mean Absolute Errors (MAEs) and Root Mean Square Errors (RMSEs) for prediction, showcasing the effectiveness of HDLA for real-time fault diagnosis in power systems.

Prediction of amplitude of fault generated travelling wave in medium-voltage grids for fault type classification

This paper proposes a new solution for the classification of short-circuits in medium-voltage distribution grids. The solution requires the measurements of the phase voltages at the time of the short-circuit and the voltage amplitudes of the short-circuit wave resulting from the fault. The values of the voltage before the short-circuit are used to calculate the expected values of the short-circuit wave amplitudes for the different types of short circuits. These calculations are based on an analysis of the charge flow between the capacitances of the power line due to the disturbance. The short-time matrix pencil method was used to detect the incoming short-circuit waves. Tests of the algorithm were carried out on the IEEE 34-bus Feeder model, taking into account the influence of voltage sensors on the measured signal values. The classification algorithm uses only non-zero components due to their relatively low attenuation. Despite that, it achieves high performance in the classification of single- and two-phase short-circuits.

Research on bearing fault diagnosis technology based on machine learning

As industrial equipment complexity continues to rise, the importance of bearings within these systems has become more critical, given their pivotal role in equipment functionality. Bearing faults can result in severe production accidents and safety issues. Hence, there is an urgent need for advanced bearing fault diagnosis technology. This study concentrates on rolling bearings, analyzing their structural characteristics and key parameters to classify fault types—inner race faults, rolling element faults, and outer race faults. Utilizing a dataset of 80 sets of bearing factory data, time and frequency domain analyses are conducted, establishing seven feature parameters (five in the time domain and two in the frequency domain). This data is organized into a 7-dimensional matrix for subsequent analysis and model development. The K-Means algorithm is chosen for its effectiveness in automatically recognizing fault patterns in rolling bearings. Training on the 7-dimensional matrix identifies four clustering centers corresponding to normal conditions, inner race faults, rolling element faults, and outer race faults. The fault diagnosis system is implemented using Python, and algorithm optimization improves efficiency. The study concludes with insights drawn from the analysis and proposes optimization methods, which contributing to advancing bearing fault diagnosis technology, particularly addressing industrial equipment reliability and safety concerns.

Gearbox faults severity classification using Poincaré plots of acoustic emission signals

Classification of fault severity in gearboxes using Acoustic Emission (AE) signals is challenging because such signals represent a highly non-linear and possibly chaotic system. Due to the common assumption of linearity, the statistical features extracted from these systems are suboptimal for the classification of fault severity. Hence, this paper uses the Poincaré plot (PP) of Acoustic Emission (AE) signals to extract useful features to classify fault type and severity in gearboxes. For this development, four fault types were applied over different gears and then tested on an experimental condition monitoring bench: broken tooth, pitting, scuffing, and cracks, each with nine severity levels. Then, the feature set was extracted from the conventional 2-D PP, composed of shape-related features and a set of features known as complex correlation measurements (CCM). The fault type and severity classification was performed using four frequency bands. Low and band-pass filtered signals obtained the highest classification accuracy with Random Forest (RF): the fault type was classified with an accuracy of 99.69%, the severity was classified with an accuracy depending on the fault type corresponding to pitting 98.76%, cracks 98.71%, broken tooth 98.96%, and scuffing 98.51%. The PP features set has a low computational cost even for large datasets representing AE signals, which can benefit the practical possibility of the implementation with high classification accuracy for different types of fault and severity levels in a gearbox.

A Decision Tree Based Ultra-high-speed Protection Scheme for Meshed MMC-MTDC Grids with Hybrid Lines

The reliable operation of modular multi-level converter based multi-terminal high voltage direct current (MMC-MTDC) grids requires high-speed and selective DC line protection. A single-ended protection technique is proposed for MTDC grids with offshore wind farms and hybrid lines that involve overhead lines and submarine cables in series. Positive and negative poles’ currents at one end of each line section are analyzed using discrete wavelet analysis (DWA). To classify fault type and identify fault zone, a fine decision tree is supplied with an energy index and the envelope slop of detail 1 coefficient (D1) obtained from DWA. Only 0.2 ms following the fault inception are needed for calculating the energy index and the envelope slope to perform relay functions. The approach was tested on a three-terminal two-poles ±\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\pm$$\\end{document} 400 kV MMC-MTDC model. The simulation results validate the effectiveness of the suggested protection technique under various fault scenarios, even with up to 200 Ω fault resistance. The proposed technique is not only able to detect the faulty line, but also it identifies overhead line faults and submarine cable faults for hybrid type lines. Moreover, the proposed technique is not affected by wind power injection changes, AC faults, or data noise. The simulated model was produced in the MATLAB/Simulink platform.

An orbit-based encoder–forecaster deep learning method for condition monitoring of large turbomachines

Most data-driven methods used in condition monitoring and early faults alarming of large turbomachines need to manually extract features for modeling, which inevitably introduces subjectivity and uncertainties. In addition, conventional deep learning methods are able to automatically classify fault types using shaft orbit images constructed from the vibration data, but image-based deep learning methods have yet been investigated for the long-term condition monitoring of a turbomachine. This study presented a novel image-based encoder–forecaster deep learning model, OrbitLSTM, to address these issues by modeling the dynamic behaviors of orbit image sequences for condition monitoring of turbomachines. The model is constructed and trained by using orbit image series in the healthy operation status through convolutional long short-term memory network. Then the temporal evolution of orbit shapes is predicted by the trained OrbitLSTM model, and used as a reference to evaluate the machine status under unknown conditions. The early fault warning is triggered when the mean square error or the structural similarity index between measurements and predictions exceeds the predefined threshold. The proposed model seamlessly combines the CNN’s capability of automatic feature extraction from images and the LSTM’s advantage of long-term memory in historical sequence behaviors. The capability and accuracy of the proposed model for fault monitoring and early warning are demonstrated using both the simulation data and the raw vibration signals collected from sensors installed in different real-world industrial turbomachines. This study provides a new image-based deep learning method for automatic condition monitoring towards the predictive maintenance of large turbomachines in various industrial fields.

Development of an Operational Digital Twin of a Locomotive Parking Brake for Fault Diagnosis

In recent years, a growing role in digital technologies has been filled by model-based digital twinning. A digital twin produces a one-to-one mapping of a physical structure, operating in the digital domain. Combined with sensor technology and analytics, a digital twin can provide enhanced monitoring, diagnostic, and optimization capabilities. This research harnesses the significant capabilities of digital twining for the unmitigated challenge of fault type classification of a locomotive parking brake. We develop a digital twin of the locomotive parking brake and suggest a method for fault type classification based on the digital twin. The diagnostic ability of the method is demonstrated on a large experimental dataset.

An unsupervised online anomaly detection method for metal additive manufacturing processes via a statistical time-frequency domain algorithm

Anomalies often occur in metal additive manufacturing from processing inconsistencies and uncertainty. A robust fault detection system that uses sensor measurements such as melt pool imaging has the potential to improve part quality and save production time by anticipating print failure. Toward this goal, we develop and validate a fault detection technique using melt pool geometry-related measurements from an in situ near-infrared optical camera. This method is unsupervised and is trained on a small dataset, mitigating human error in classifying fault types, and reducing lead times for preparing training datasets. Furthermore, this method uses learned geometry-informed nominal behavior of the melt pool signal to make informed decisions on the process health. There are spatial-temporal characteristics embedded in the melt pool images, caused by the periodicity in the geometry-dependent raster pattern. These characteristics can be captured in the frequency domain using the signal spectrogram, a representation of the frequency content over time. Defects will appear in the spectrogram, disrupting the healthy spectral response. To quantify healthy spectrograms, we use principal component (PC) decomposition to extract the features of these spectrograms as a set of nominal basis vectors. Anomaly detection is then performed by calculating the error between the original and reconstructed spectrogram vector by projection of the spectrogram PCs onto the nominal basis. The reconstruction error for anomalous signals is larger than that from healthy signals, which is then used for fault detection. A one-tailed statistical test is used to determine the fault detection threshold for the reconstruction error signal. This method is tested on three raster patterns and performs better than a comparative time-series thresholding method. We demonstrate that this time-frequency algorithm can detect both temporal faults (which occur at a single time instant) and spatial faults (such as those introduced by an improper sintering), differentiating them from nominal operation.

Development of an Operational Digital Twin of a Locomotive Braking System Solenoid Valve for Fault Classification

In recent years, a growing role in digital technologies has been filled by model-based digital twinning. A digital twin produces a mapping of a physical structure, operating in the digital domain. Combined with sensor technology and analytics, a digital twin can provide enhanced monitoring, diagnostic, and optimization capabilities. This research harnesses the significant capabilities of digital twining for the unmitigated challenge of fault type classification of a locomotive braking system solenoid valve. We develop a digital twin of the solenoid valve and suggest a method for fault type classification based on the digital twin. The diagnostic ability of the approach is demonstrated on a large experimental dataset.

Improved fault diagnosis method of electric gate valve in nuclear power plant

The vibration signal (VS) analysis and acoustic emission signal (AES) analysis methods are the common effective diagnostic methods in valve fault diagnosis (FD)(Classify fault types). However, these two methods are disadvantaged by the surrounding environment and noise signals produced during acquisition. To obtain an improved FD effect of an electric gate valve (EGV) in nuclear power plant, the two previous methods are ameliorated by a noise reduction processing, then the signal feature extraction is performed. The FD and degree evaluation are carried out for the common fault types of the EGV taken as object. In this paper, three-phase unbalanced faults (Acquisition of VS), packing damage faults (Acquisition of VS) and internal leakage faults (Acquisition of AES) are taken as examples. The Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) algorithm composed with Fast Fourier Transform (FFT) and the SAEU3H series digital acoustic emission detection system (SAEU3H SDAEDS) are used to noise reduction and extraction of parameters of the VS and AES, respectively. Then, the extracted feature parameters are brought into the Bi-directional Long Short-Term Memory (Bi-LSTM) deep neural network for FD and fault degree evaluation (Assess the severity of failure occurrences). The experimental results show that the method adopted in this paper has high FD accuracy and low fault evaluation error.

A Deep Learning Method for Rolling Bearing Fault Diagnosis Based on Attention Mechanism and Graham Angle Field.

Focusing on the low accuracy and timeliness of traditional fault diagnosis methods for rolling bearings which combine massive amounts of data, a fault diagnosis method for rolling bearings based on Gramian angular field (GAF) coding technology and an improved ResNet50 model is proposed. Using the Graham angle field technology to recode the one-dimensional vibration signal into a two-dimensional feature image, using the two-dimensional feature image as the input for the model, combined with the advantages of the ResNet algorithm in image feature extraction and classification recognition, we realized automatic feature extraction and fault diagnosis, and, finally, achieved the classification of different fault types. In order to verify the effectiveness of the method, the rolling bearing data of Casey Reserve University are selected for verification, and compared with other commonly used intelligent algorithms, the results show that the proposed method has a higher classification accuracy and better timeliness than other intelligent algorithms.

분산형 전원이 연결된 배전계통에서의 분리된 상별 인공지능 모델을 이용한 고장종류 및 고장구간 판별 방법

In recent years, the number of distributed generation (DG) in distribution network (DN) has increased. The increasing proportion of DG in DN brings benefits through improved network reliability and reduced transmission power losses. However the connection of DG makes network be more complicated, it changes the unidirectional flow of currents and power to bidirectional flow and thus protection based on over-current relay has limitations of protection coordination. These limitations may occur malfunction or misbehavior of protective devices so it requires new protection methods. This paper proposes a new method using artificial neural network (ANN) can adjust on DN with DG to allocate fault section and to classify fault type as part of solving limitations of protection coordination. This ANN model is separated to each phase and use magnitude and phasor of voltages and currents extracted from each generation sides. The new proposed method is applied to unbalanced distribution network model and verified to be useful in DN through computer simulations.

Shunt faults detection and classification in electrical power transmission line systems based on artificial neural networks

PurposeThis study aims to focus on an adaptive method for fault detection and classification of fault types that trigger in three-phase transmission lines using artificial neural networks (ANNs). The proposed scheme can detect and classify several types of faults, including line-to-ground, line-to-line, double-line-to-ground, triple-line and triple-line-to-ground faults.Design/methodology/approachThe fundamental components of three-phase current and voltage were used as inputs in the ANNs. An analysis of the impact of variations in the fault resistance, fault type and fault inception time was conducted to evaluate the ANNs performance. The survey compares the performance of the multi-layer perceptron neural network (MLPNN) and Elman recurrent neural network trained with the backpropagation learning technique to improve each of the three phases of the fault detection and classification process. A detailed analysis validates the choice of the ANNs architecture based on the variation in the number of hidden neurons in each step.FindingsThe mean square error, root mean square error, mean absolute error and linear regression are measured to improve the efficiency of the ANN models for both fault detection and classification. The results indicate that the MLPNN can detect and classify faults with a satisfactory performance.Originality/valueThe smart adaptive scheme is fast and accurate for fault detection and classification in a single circuit transmission line when faced with different conditions and can be useful for transmission line protection schemes.

Deep learning-based application for fault location identification and type classification in active distribution grids

The high penetration of distributed energy resources, especially weather-dependent sources, even at the edge of the distribution grids, has increased the power system uncertainties and drastically shifted the operational status quo for the system operators. For the operators to ensure the uninterrupted electricity supply of the end-consumers, the fast and accurate response to fault events is of critical importance. This paper proposes a data-driven fault location identification and types classification application based on the continuous wavelet transformation and convolutional neural networks optimally configured through Bayesian optimization. This application leverages the proliferation of high-resolution measurement devices in distribution networks. It can locate the exact position of the short-circuit faults and classify them into eleven different types. Its intrinsic models grasp the spatial characteristics and the converted in frequency domain temporal ones of the three-phase voltage and current timeseries measurements stemming from the field devices, thus increasing the operators’ visibility of their networks in real-time. We conduct simulations through synthetic data, which we provide in an open-source repository, that replicate a wide range of fault occurrence scenarios with eleven different types, with the resistance ranging from 50Ω to 2kΩ and with duration from 20ms to approximately 2s, under noise conditions injected by devices and load variability. The results showcase the efficacy of the proposed method reaching an accuracy of 91.4% for fault detection, 93.77% for correct branch identification, 94.93% for fault type classification, and RMSE value of 2.45% for location calculation.

Phasor measurement unit application-based fault allocation and fault classification

<span>This paper makes a contribution to the field of fault location finding in a new way that helps in the improvement of grid reliability. This paper proposes a study-based approach for fault allocation and fault type classification that uses the study of voltage and current frequency during the abnormal condition. Although, ideally frequency of voltage and current are the same in the abnormal condition they may differ from each other. This difference in frequency is separately measured by the phasor measurement unit (PMU) block at MATLAB/Simulink platform. The PMU (PLL-based, positive-sequence) block is inspired by the IEEE Std C37.118.1-2011. In this approach, we measure the line voltage and current frequency variation with the help of installed PMU after this we present this measurement in characteristics form with the help of the scoping tool in MATLAB/Simulink and study them one by one, and proposed a conclusion for fault location identification and fault type classification. The proposed approach is able to identify the source side and load side fault location and also able to classify faults into two categories namely symmetrical and asymmetrical. The proposed approach is tested on two MATLAB/Simulink models and observed satisfactory.</span>

Rotor Fault Diagnosis Method Using CNN-Based Transfer Learning with 2D Sound Spectrogram Analysis

This study discusses a failure detection algorithm that uses frequency analysis and artificial intelligence to determine whether a rotor used in an industrial setting has failed. A rotor is a standard component widely used in industrial sites, and continuous friction and corrosion frequently result in motor and bearing failures. As workers inspecting failure directly are at risk of serious accidents, an automated environment that can operate unmanned and a system for accurate failure determination are required. This study proposes an algorithm to detect faults by introducing convolutional neural networks (CNNs) after converting the fault sound from the rotor into a spectrogram through STFT analysis and visually processing it. A binary classifier for distinguishing between normal and failure states was added to the output part of the neural network structure used, which was based on the transfer learning methodology. We mounted the proposed structure on a designed embedded system to conduct performance discrimination experiments and analyze various outcome indicators using real-world fault data from various situations. The analysis revealed that failure could be detected in response to various normal and fault sounds of the field system and that both training and validation accuracy were greater than 99%. We further intend to investigate artificial intelligence algorithms that train and learn by classifying fault types into early, middle, and late stages to identify more specific faults.

Power quality improvement in power electronics systems using machine learning

This study focuses on improving power electronic components and devices, which have gained popularity due to their compact size and precise control over the output voltage and current. They are widely used in renewable energy systems, such as converters and inverters, and in industrial drives as control devices. However, the switching process in power electronic components introduces nonlinear behavior, leading to the generation of harmonics and power quality problems. To address these issues, various methods and techniques recommended by industry standards have been proposed to mitigate or eliminate harmonics and ensure the power quality. These standards provide limits and guidelines for assisting customers and manufacturers in maintaining acceptable power quality levels. Machine-learning algorithms offer a promising approach for improving power quality by leveraging data from a system to make predictions or classifications. In the context of power electronics, machine-learning algorithms can be trained to classify fault types, identify the exact location of faults, predict the remaining life of power electronic components, and detect voltage disturbances. Different machine learning techniques can be employed, depending on the specific application. For example, fault classification and fault location identification can be achieved through supervised learning algorithms, whereas predicting component life and detecting voltage disturbances can utilize techniques such as regression or anomaly detection. By leveraging the power of machine learning, enhanced reliability, performance, and efficiency can be achieved in various applications, thereby contributing to the overall advancement of power electronics technology.

Pulse Bag Filter Fault Classifier Research and Analysis

Aiming at the problem of fault alarm flooding, a fault classifier is established to automatically classify fault types in combination with unsupervised learning methods in machine learning. The k-means method is used to classify the alarm types in dimensionality reduction, and the optimal K value is selected as 11 through various evaluation indexes, and the faults are added to the new category labels according to the center of each cluster after clustering. The random forest and support vector classification algorithm is used to establish a fault classifier model on the fault alarm dataset with new labels, and the unified analysis of the fault type prediction results of the test set unlabeled fault data set is compared with the random forest algorithm, support vector classification, and k-means method. The results show that the unity rate of the prediction results of the three types of classifiers on the test set when taking different K values can reach up to 67%, and the uniformity rate of the prediction results of the three models exceeds 0.998, which has a high degree of uniformity, indicating that the establishment of the fault alarm classifier for the pulse bag dust collector is feasible and has good classification accuracy, which can help maintenance personnel quickly locate and troubleshoot faults, and improve the intelligence level of equipment.