Fault Current Signals Research Articles

Application of symmetric uncertainty and emperor penguin–grey wolf optimisation for feature selection in motor fault classification

AbstractThe authors present a model for diagnosing motor faults based on machine learning, demonstrating advantages over other algorithms in terms of both improved fitness values and reduced running time. The structure of the model involves three primary phases: feature extraction, feature selection and classification. During the feature extraction phase, crucial features are identified using empirical mode decomposition, fast Fourier transform and multiresolution analysis, resulting in a total of 144 features. The feature selection stage employs a new strategy that combines symmetrical uncertainty in the filter approach with the binary grey wolf optimiser and emperor penguin optimiser in the wrapper approach. Finally, a support vector machine is used for classification to generate fitness values. To validate the model's effectiveness and accuracy, motor fault current signal datasets, case Western Reserve University (CWRU) benchmark datasets and mechanical failure prevention technology benchmark datasets are utilised. In the motor fault current signal dataset, the highest average accuracy achieved is 99.95%, with a minimum average running time of 88.02 s obtained under ∞dB conditions. Regarding benchmark datasets and mechanical failures at CWRU, using the prevention technology benchmark dataset resulted in classification accuracies of 99.54% and 99.52%, respectively. Comparative analysis with traditional algorithms reveals that symmetric uncertainty and emperor penguin–grey wolf optimisation model outperforms traditional models in terms of performance.

Fault detection through discrete wavelet transforms and radial basis function neural network in shunt compensated distribution systems

This paper discusses a critical study of fault detection, classification, and identification in distribution systems compensated through a distribution static synchronous compensator (D-STATCOM) using discrete wavelet transforms (DWT) and a radial basis function neural network (RBFNN). The efficacy of the proportional integral (PI) controller is discussed in the simulation during normal, voltage sag, and swell conditions. Then, the Daubechies mother wavelet (db4) is used here to extract and decompose the fault current signals because of its high accuracy of detection with less processing time. The models are subjected to different conditions of faults, such as line to ground (LG), line to line (LL), double line to ground (LLG), and three phases (LLL and LLLG) with different fault resistance. The novelty of the scheme is that DWT and RBFNN techniques were compared and proven effective in demonstrating the faulty conditions. To validate the proposed approach, a simulation study is carried out in MATLAB with various operating conditions, and it is shown that the proposed method can depend on abundantly protecting the distribution grid from faulty conditions.

Research on the longitudinal protection of a through-type cophase traction direct power supply system based on the empirical wavelet transform

This paper proposes a longitudinal protection scheme utilizing empirical wavelet transform (EWT) for a through- type cophase traction direct power supply system, where both sides of a traction network line exhibit a distinctive boundary structure. This approach capitalizes on the boundary’s capacity to attenuate the high-frequency component of fault signals, resulting in a variation in the high-frequency transient energy ratio when faults occur inside or outside the line. During internal line faults, the high-frequency transient energy at the checkpoints located at both ends surpasses that of its neighboring lines. Conversely, for faults external to the line, the energy is lower compared to adjacent lines. EWT is employed to decompose the collected fault current signals, allowing access to the high-frequency transient energy. The longitudinal protection for the traction network line is established based on disparities between both ends of the traction network line and the high-frequency transient energy on either side of the boundary. Moreover, simulation verification through experimental results demonstrates the effectiveness of the proposed protection scheme across various initial fault angles, distances to faults, and fault transition resistances.

Deep learning-based integrated attack detection framework to protect distance relays against cyberattacks

The cyber security of communication-assisted intelligent relays at the IEC61850-enabled modern digital substations has become a prime concern among power companies and regulatory bodies. Particularly, distance relay IEDs are easy to manipulate to send false trip command to circuit breakers through random false data injection and replay attacks. With these attacks evolving and becoming more sophisticated, existing intrusion detection techniques can be easily compromised, resulting in unwanted line outages. This work proposes a convolutional-based autoencoder and Siamese neural network deep learning integrated attack detection framework (IADF) to differentiate these attacks from actual faults. The framework works in conjunction with distance relay IEDs to issue genuine trip commands to circuit breakers. The autoencoder, trained on fault voltage and current signals, detects the false data injection attacks using reconstruction error, while the siamese neural network trained on similar and dissimilar pairs of the fault voltage and current signals detects the fully synchronized replay attacks using similarity estimation with fault data stored on database. The validity of the proposed framework is assessed using IEEE 39 and 118 bus test system simulated on PSCAD/EMTDC software. The validation results suggest that the proposed IADF can accurately discriminate the faults from attacks and is resilient to various system parameter changes.

A Novel Method for Removal of Dual Decaying DC Offsets to Enhance Discrete Fourier Transform-Based Phasor Estimation

This paper presents a novel method for the removal of dual decaying DC offsets (DDCOs) to enhance discrete Fourier transform-based phasor estimation. The method proposed in this paper uses the sum of even samples and the sum of odd samples from the input signal to remove the AC components, thereby precisely estimating the primary and secondary DDCOs. The fluctuations induced by DCCOs present in the output of traditional DFT methods are eliminated by using the estimated DCCOs. The performance of the method was evaluated in terms of both mathematical-generated signals and fault current signals from a 154 kV Korean overhead transmission system. PSCAD/EMTDC was used to generate fault current signals at various fault distances and angles. The results show that the proposed method can estimate the phasor of the fundamental frequency component accurately regardless of the primary and secondary DCCOs. The paper concludes by describing the hardware implementation in a prototype unit that considers the real-world requirements of power system protection.

Stator-rotor fault diagnosis of induction motor based on time-frequency domain feature extraction

Since the induction motor operates in a complex environment, making the stator and rotor of the motor susceptible to damage, which would have significant impact on the whole system, efficient diagnostic methods are necessary to minimize the risk of failure. However, traditional fault diagnosis methods have limited applicability and accuracy in diagnosing various types of stator and rotor faults. To address this issue, this paper proposes a stator-rotor fault diagnosis model based on time-frequency domain feature extraction and Extreme Learning Machine (ELM) optimized with Golden Jackal Optimization (GJO) to achieve highprecision diagnosis of motor faults. The proposed method first establishes a platform for acquiring induction motor stator-rotor fault data. Next, wavelet threshold denoising is used to pre-process the fault current signal data, followed by feature extraction to perform time-frequency domain eigenvalue analysis. By comparison, the impulse factor is finally adopted as the feature vector of the diagnostic model. Finally, an induction motor fault diagnosis model is constructed by using the GJO to optimize the ELM. The resulting simulations are carried out by comparing with neural networks, and the results show that the proposed GJO-ELM model has the highest diagnostic accuracy of 94.5%. This finding indicates that the proposed method outperforms traditional methods in feature learning and classification of induction motor fault diagnosis, and has certain engineering application value.

Transformer Faults Classification Based on Convolution Neural Network

This paper studies the latest advances made in Deep Learning (DL) methods utilized for transformer inrush and fault currents classification. Inrush and fault currents at different operating conditions, initial flux and fault type are simulated. This paper presents a technique for the classification of power transformer faults which is based on a DL method called convolutional neural network (CNN) and compares it with traditional artificial neural network (ANN) and other techniques. The inrush and fault current signals of the transformer are simulated within MATLAB by using Fourier analyzers that provides the 2nd harmonic signal. The 2nd harmonic peak and variance statistic values of input signals of the three phases of transformer are used at different operating conditions. The resulted values are aggregated into a dataset to be used as an input for the CNN model, then training and testing the CNN model is performed. Consequently, it is obvious that the CNN algorithm achieves a better performance compared to other algorithms. This study helps with easy discrimination between normal signals and faulty signals and to determine the type of the fault to clear it easily.

Mathematical Morphology-Based Fault Detection in Radial DC Microgrids Considering Fault Current From VSC

The paper proposes a fast fault detection method for radial DC microgrids established on mathematical morphology (MM) denoising filters and detection principles utilizing only local measurements. The proposed fault detection algorithm utilizes the DC fault current signal from a Voltage Source Converter (VSC) and detects a fault. Problems related to communication delays are thus avoided, while preserving the low cost and computational burden. The proposed fault detection framework includes five different MM-based denoising filters, employed as pre-signal processing methods whose features are assessed and compared. The influence of selecting an appropriate type and length of a structuring element (SE) on the proposed method is analyzed. The developed fault detection method based on MM is used for extracting transient features and detecting pole-to-pole (PP) and pole-to-ground (PG) faults within milliseconds of their occurrence. The proposed method is demonstrated using simulation tests in MATLAB/Simulink. The results indicate that the proposed method enables reliable, accurate and fast detection of both the PP and the PG faults. Computer simulations are carried out to verify that the proposed method distinguishes faults from overload.

Application of Time-Frequency Analysis with Selection of Suitable Mother Wavelets for Transient States of HVDC Transmission Lines

Recent developments in power semiconductor devices have made High Voltage Direct Current (HVDC) transmission systems the most crucial technology for future power systems. Today several HVDC projects have been commissioned around the world. The primary concern for such an advanced HVDC grid is to follow standard grid codes to maintain the safety and stability of the electrical equipment during transient and dynamic circumstances. Although the HVDC system has been recognized as the most prominent technology to transmit electric power, some issues must be addressed, such as selective and quick detection of faults. For selective fault detection, it is essential to look into the several fault detection strategies that fortify the HVDC system with time-frequency signal processing techniques. This research uses a time frequency-based method to identify the different HVDC transmission line faults. The most suitable mother wavelet for various transient states is selected by analyzing fault voltage and current signals based on variance, standard deviation, and range parameters. An experimental setup of a thyristor-based HVDC transmission system has been developed in the MATLAB/Simulink environment, and multiple fault scenarios have been analyzed. The simulation outcomes reveal that the most effective mother wavelet for detecting DC line, symmetrical and unsymmetrical faults occurs at the 6th level of fault signal decomposition. By examining current and voltage signals, our work has determined the best mother wavelet for various faults. Using more than one mother wavelet to detect faults is more reliable than using a single mother wavelet to detect all faults. Additionally, it has been noted that each fault's voltage and current signals have two distinct mother wavelets.

Squaring and lowpass filtering data-driven technique for AC faults in AC/DC lines

Transient events that result from the incorporation of HVDC into the HVAC power transmission system make fault identification a difficult task. To minimize transient power outages, anomalies must be identified and categorized as quickly as feasible using robust schemes. In the proposed scheme, the multi-classification of AC faults in hybrid transmission lines is performed. A neural network has been employed for the correct recognition and classification of AC faults. The proposed scheme initially uses squaring and lowpass filtering techniques along with, transient energy, negative sequence of voltage, and current as features to pre-process the fault voltage and current signals. The extracted features are then used to form the neural network's input for training and testing. We performed a complete assessment study on the developed AC/DC test system employing MATLAB/Simulink software to ensure the stability and reliability of the presented technique. The technique is verified under noise-added data and compared with other schemes to ensure efficacy. The test result shows that the proposed technique has successfully classified the AC faults with an accuracy of 99.3% in AC/DC transmission lines.

Data window selection for removal of decaying DC component in current phasor estimation for distance protection

In distance protection, the fault current signal is usually distorted by the decaying direct current (DC) and harmonics preventing accurate and fast estimation of the fault current phasor. This paper presents a method based on the carefully selected data windows and least square (LSQ) method to improve the performance of the current phasor estimation. The method comprises a short and variable data window for fast operation of distance relay for close faults and a fixed data window for the accurate operation of the relay for faults close to the boundary of Zone 1. The short and variable data window can achieve high speed in the phasor estimation in a quarter of cycle, whereas the second algorithm uses a fixed data window with an excellent frequency response to increase the accuracy of the phasor estimation. The proposed method is first tested using mathematical formulation under different conditions such as different characteristics of the DC component, noise, harmonics and frequency deviation. Then the method is tested by more realistic samples taken from a simulated transmission system under different fault conditions for flexibility, speed and accuracy. All the test results validate the aforementioned features of the proposed algorithms.

VSC control strategy for HVDC compensating negative sequence and decaying DC components

This paper proposes a voltage source converter (VSC) control strategy for the high voltage direct current (HVDC) transmission systems compensating the negative sequence and decaying DC (DDC) components. The mathematical model of the grid fault current is established firstly, and the DDC component is extracted using the property that the integral of the periodic signal is 0. Secondly, the fundamental wave component of the grid fault current signal is separated in the rotating frame system to obtain the compensation current command of VSC-HVDC. Finally, the performance of the proposed control strategy is compared with the conventional inverter control strategy when compensation the negative sequence and decaying DC (DDC) components by the experimental platform, and the experimental results show that the flexible HVDC transmission system with the proposed control strategy has the desired ability of compensating the negative sequence and DDC components simultaneously.

A real-time DC faults diagnosis in a DC ring microgrid by using derivative current based optimal weighted broad learning system

This paper presents a novel approach for DC faults diagnosis in renewables based DC-ring microgrid (DC-RM). The proposed novel approach consists of a second-order derivative current (SODC) approach, which is adequate for isolation of faults due to its accurate fault detection and faster response during fault. Moreover, this approach is also used to estimate the fault location using cable parameters. On the other hand, most powerful weighted broad learning system (WBLS) is developed for faults classification. However, WBLS has a major challenge i.e., random generation of weights or fixed weights for the input data, which is connected to WBLS network. To address that and generate the suitable weights according to the statistical feature data, optimized WBLS is developed by hybridizing the sine–cosine algorithm (SCA) and chaotic salp swarm algorithm (CSSA), called SC-CSSA-WBLS. Here, the most influential features are used to extract the data from the simulated DC fault current signals. The role of SSA and SCA in the WBLS is to enhance the initialization of population, convergence speed, and search capabilities of both local exploitation and global exploration. Further, the efficacy of the proposed algorithm is validated during different case studies and its superiority is evidenced in terms of detection time, relative computational time, classification accuracy, and relative error when compared to state-of-the-art techniques in MATLAB/Simulink environment. Finally, a real-time hardware setup is implemented using dSPACE DS1104 embedded processor to isolate the various faults that occurred on the DC microgrid accurately using adaptive threshold strategy of SODC approach.

Series Arc Fault Detection Based on Dual Filtering Feature Selection and Improved Hierarchical Clustering Sensitive Component Selection

Affected by the downstream load of the line, the low-voltage AC series arc fault signal presents nonlinear, non-stationary and random characteristics, which bring great difficulties to the extraction of arc fault features and lead to low arc detection accuracy. To solve this problem, a low-voltage AC series arc fault detection method based on strong discriminative features and sensitive components is proposed. In this work, firstly, the arc fault current signal is collected based on the self-built experimental platform, and the arc fault current components are obtained by the complete ensemble empirical mode decomposition with the adaptive noise (CEEMDAN) method. Secondly, 16 feature indexes based on time window are used to describe the component signals, and an effective feature selection method based on the maximum mutual information coefficient and the significance of feature changes is proposed to realize the mining of highly identifiable features. Then, an improved hierarchical clustering algorithm and a sensitive component selection strategy combining the saliency of feature changes are proposed to eliminate the interference of redundant components. Finally, a strong discriminative feature library of sensitive components of current signal is constructed, and series arc fault detection is realized based on support vector machine. Experimental results show that the proposed method is feasible and effective in arc current feature extraction and fault detection, and provides a reference for low-voltage AC series arc fault detection.

Research Based on Improved CNN-SVM Fault Diagnosis of V2G Charging Pile

With the increasing number of electric vehicles, V2G (vehicle to grid) charging piles which can realize the two-way flow of vehicle and electricity have been put into the market on a large scale, and the fault maintenance of charging piles has gradually become a problem. Aiming at the problems that convolutional neural networks (CNN) are easy to overfit and the low localization accuracy in fault diagnosis of V2G charging piles, an improved fault classification model based on convolutional neural networks (CNN-SVM) is proposed. Firstly, the hardware adaptation optimization is carried out for the CNN structure, the wavelet packet transformation is used to extract the fault current signal feature information into the CNN, and the CNN-SVM model is constructed by SVM (Support Vector Machine) instead of the SoftMax classifier in the CNN. The PSO (particle swarm algorithm) is used to optimize the parameters of the SVM model to obtain the optimal model. Finally, the superiority of the proposed method is verified by multi-working cases. The experimental results show that the fault classification accuracy of the CNN-SVM model is far higher than that of the traditional deep learning network and has practical significance for fault diagnosis of the switch module of the charging pile.

A Hybrid Approach for Low-Voltage AC Series Arc Fault Detection

In a low-voltage electric distribution network, arc fault presents a high energy density electricity-discharging phenomenon between conductors, which is often caused by aging of electric facilities, loose contacts and terminals, or insulation failure due to internal and external destructions. A large amount of heat may be created during this discharging, which will further cause the risk of fire hazards to mitigate in the residential environment. Currently, many utility grid operators and electricity users are still devoted to seeking effective detection technology for arc fault protection. This paper proposes a hybrid approach that combines discrete wavelet transform (DWT), empirical mode decomposition (EMD), and dynamic time warping (DTW) methods for low-voltage AC series arc fault detection. In DWT, it uses time–frequency domain characteristics of the arc current signal to extract the occurrence of arc fault. In EMD, it decomposes the complex arc fault current signal into a finite intrinsic mode (IMF) signal; then, instantaneous amplitude of IMF signal is obtained by Hilbert–Huang transform (HHT) as a feature for arc fault identification. Firstly, the results of arc fault detections depend on the results from DWT and EMD. When both two methods detect different results, DTW method will be activated, using the similarity measurements between normal and arc fault current waveforms as an assistant measure to determine the occurrence of arc fault. The performance of the proposed approach is tested and validated using various electric appliances, and the results show that the proposed approach can effectively detect low-voltage AC series arc fault.

Teager energy assisted variational mode decomposition‐based fault location technique for STATCOM compensated system

AbstractModern power systems are structurally complex and thus experience transient events. For secure and reliable operation, quick system restoration is required whenever transient phenomenon like line faults occurs. Fault location techniques (FLTs) aid in faster restoration process. However, the accuracy of FLTs gets affected in presence of static synchronous compensator (STATCOM) since STATCOM compensates the dip in measured voltage during fault thereby affecting the fault current signals. Therefore, in this paper, a Teager energy assisted variational mode decomposition (VMD) based traveling wave FLT is developed for transmission lines with STATCOM. The proposed FLT considers two‐terminal instantaneous voltage signals for locating faults. In this FLT, the signals are first decoupled into modal signals and then aerial mode of the signals is chosen for decomposition using VMD. The signals are decomposed via VMD to level three producing mode three signals. Teager energy of the resultant signals is obtained and from the first peak of the energy signals, arrival time of waves (ATWs) is estimated. These estimated ATWs are used to calculate fault location and percentage error. To validate the performance of proposed FLT, several test cases with different fault scenarios are considered and the results are compared with existing techniques. The results prove efficacy of proposed FLT for studied system.

A reconstruction based adaptive fault detection scheme for distribution system containing AC microgrid

The occurrence of a single line to ground fault, including high impedance faults (HIFs), in various grounding systems,results in a low magnitude fault current. Due to this, the present protection device has trouble finding the fault and faulty feeder. This manuscript attempts to propose a new fault detection strategy based on a sliding window-based recursive matrix pencil method to address this problem. The suggested technique starts by utilizingrecursive matrix pencil method to reconstruct the fault current signal, and after that, it calculates the transient error by deducting the reconstructed fault current signal from the original signal. This error serves as a fault detection index, which is also utilized to categorize the defective feeder. The suggested approach does not call for the computation of any zero sequence components and does not entail any extra voltage-based criteria, in contrast to existing techniques. For three different distribution networks with AC microgrids that were modeled using EMTDC/PSCAD 5.0, the superiority of the given technique was tested. Results, which were acquired by simulating several worst-case scenarios, demonstrate the effectiveness of the strategy being discussed (fault and non-fault). Additionally, the applicability of the proposed protective mechanism is supported by the results of two field test cases, which validate its use in actual situations.

VSC-HVDC Fault Location Method Based on Effective Feature and Depth Belief Network

Aiming at the problems that most fault location methods are applied to the high voltage direct current (HVDC) transmission system, such as low location accuracy and vulnerability to noise, a voltage source converter (VSC)-HVDC fault location method based on effective feature and deep belief network is proposed. Firstly, the Karenbauer phase mode transform is used to decouple the original fault current data to eliminate the coupling effect between positive and negative lines. Then, Hilbert Huang (HH) transform is used to obtain six effective features of information from frequency domain and time domain analysis to fully characterize the system fault current signal. Finally, the extracted effective features are input into the depth belief network for learning, so as to estimate the fault location in the transmission line. VSC-HVDC system is built based on the MATLAB platform for experimental analysis. The results show that the fault location obtained by the proposed method has the smallest overall fault location error, and the maximum and minimum errors are divided into about 0.52 km and -0.95 km. Compared with other methods, it has certain advantages in fault location.

High-impedance fault location method of three-phase distribution lines based on electromagnetic time reversal

• Expanding the application scenario of the criterion of the minimum mirrored FCSE. • Accurately locating the single-phase HIF and identifying the fault phase. • Not affected by noise, GPS time synchronization, and the time delay of the device. • Location and diagnosis of single-phase HIF, phase to phase HIF, and three-phase HIF. Accurate location of high-resistance grounding faults in distribution lines is of great significance for improving the power supply reliability of power systems. In this study, a simulation model of a 10 kV three-phase distribution line is built using a numerical simulation platform. Further, a fault location method of three-phase distribution lines is proposed on the basis of the criterion of the minimum mirrored fault current signal energy (FCSE). The positioning accuracy of the high-impedance fault (HIF) is studied under different influencing factors. The results show that the proposed method can accurately locate the single-phase HIF of the three-phase distribution line, and the positioning error is within 50 m. However, it can only locate the fault when the fault resistance is 5 kΩ and below. In addition, the positioning accuracy of the proposed method for the HIF is not affected by noise and the sampling frequency, although the location accuracy under different fault inception angles is affected by fault resistance. Further, when the GPS time synchronization error is within 2 μs or the time delay of the transient signal device is within 2 ms, accurate positioning of the single-phase HIF can be realized. Moreover, by analyzing the three-phase FCSE distribution calculated in the reversed time, positioning and diagnosis of three HIF types, namely the single-phase grounding fault, phase-to-phase fault, and three-phase grounding fault, can be realized. Finally, fault phase identification of the single-phase HIF can also be achieved.