Grid Fault Detection Research Articles

Integrating Machine Learning and Feature Extraction for Islanding Detection in Grid-Connected Photovoltaic Systems: A Hybrid Intelligent Approacha

Introduction: In recent years, the integration of renewable energy sources, particularly Photovoltaic (PV) systems, into the grid has garnered considerable attention. However, the distributed nature of these grid-integrated PV systems has introduced challenges concerning grid faults and maintenance. Method: This paper aims to present a pioneering approach to augment the monitoring of gridintegrated PV systems by integrating intelligent methods, specifically machine learning and feature extraction techniques. The primary focus of this approach is on islanding detection, which involves promptly identifying grid faults or maintenance challenges and initiating the grid's transition to an isolated mode of operation. To accomplish this, an intelligent signaling method is employed, capitalizing on the capabilities of Distributed Generation (DG) networks. By recording critical signals such as voltage, current, and frequency at common coupling points, fault conditions can be accurately detected. Signals obtained from the grid are subjected to wavelet transformation to extract pertinent information that characterizes fault conditions. These extracted features are then utilized as inputs to machine learning methods, facilitating the proposal of intelligent islanding scenarios. To assess the efficacy of the proposed approach, simulations are conducted on a grid-connected PV system. The recorded signals are meticulously analyzed, and the extracted features are employed to train machine learning models. The performance of these models is evaluated based on their ability to detect fault conditions and initiate appropriate islanding scenarios accurately. Result: The results obtained demonstrate the immense potential of the proposed approach in bolstering the monitoring of grid-integrated PV systems. Conclusion: By synergizing machine learning techniques with feature extraction and intelligent signaling methods with the KNN-Confusion Matrix, all predicted labels match the true labels, resulting in a 100% accuracy. The detection of grid faults and maintenance challenges can be substantially improved, thereby fostering more efficient and dependable operation of these systems.

A customised artificial neural network for power distribution system fault detection

AbstractMachine learned fault detection approaches are being increasingly used for fault detection in distribution grid. However, the performance of the models can be improved by customizing the models. In this regard, a customised artificial neural network (CANN) for fault detection in a distribution grid is proposed in this paper. The proposed work develops a CANN that combines the “up‐pyramid” and “down‐pyramid” model of ANN into a “custom‐pyramid” model. As a result, the same model can be used both for determining the types of fault as well as its location. The data needed to train the model has been taken from a reconfigured IEEE‐33 bus distribution system developed in Typhoon HIL real‐time simulator. Spectral‐kurtosis is utilized for extraction of features of the faulted transient signals which are used as input data to develop the CANN. The result showcases that the reduction of input features reduces computational complexity without compromising its accuracy. The proposed model classifies fault location with an accuracy of 95.43%. The proposed method also identifies fault type with an accuracy of 96.08%. Several test cases have been developed to test the method. The method proved to be able to perform in most of the cases.

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

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

A wavelet based PSD approach for fault detection and classification in grid connected inverter interfaced microgrid

The recent progress in the area of microgrid protection has led to the application of data-driven and signal processing techniques for fault detection in microgrids. The deployment of inverter interfaced distributed generators makes the traditional protection schemes ineffective. In this paper a wavelet based power spectral density (PSD) has been utilized to detect fault, identify faulty phases and locate fault in the microgrid. Wavelet based PSD decouples the frequency information from the time information and it is the optimal technique to reveal the change of distribution of harmonics in non-stationary faulty signals as for fault analysis it can consider the faulty frequency band instead of one particular harmonic. The proposed technique can detect fault and identify faulty phases by analyzing current signals from the substation only. The normalized value of PSD of three phase current signals has been considered for fault detection and probabilistic distribution of PSD has been considered for faulty phase identification. The normalized value of PSD of the current signals collected from both ends of the main feeders and one end of the non-DG sub-feeder have been used to locate the faulty feeder and sub-feeder. Appropriate threshold values have been considered to detect fault and to locate the faulty feeder. To locate faults, the technique does not require synchronized signals and thus provides a promising and economical solution for inverter interfaced microgrid. It does not require huge amount of data set for training and testing as it is essentially a threshold based technique and also does not require collection of data at high sampling rate. The proposed scheme has been extensively tested for high impedance fault (HIF), unbalanced loading, load switching, variation of DG parameters, measurement uncertainty and presence of noise in the signal with signal to noise ratio up to 10 dB. This technique provides accurate results for fault resistance from 0 Ω to 300 Ω and up to 10% measurement error. The performance of the proposed technique has been verified for feeder length variation up to 20%. Half cycle of current signal only after fault is required and three levels of wavelet decomposition has been considered. The simulation results and comparative assessment demonstrate the efficacy of the proposed scheme.

Soft computing based smart grid fault detection using computerised data analysis with fuzzy machine learning model

Electrical grids are more dependable, secure, and significant smart grid (SG) technologies. For effective and dependable electricity distribution, new risks are raised by its high reliance on digital communication technologies. The best grid monitoring and control skills are essential for system reliability. Among other things, SG applications include three key challenges: managing big data volumes, having enough real-time capable measurement instruments, and two-way low-latency communication. This study proposes a unique method for detecting faults in the smart grid via the use of data monitoring and classification using a fuzzy machine learning model. Here, enhanced smart sensor metering performed in the cloud at the network's edge has been used to track data from the smart grid. Fuzzy reinforcement encoder adversarial NN has then been used to categorise the tracked data. Experimental analysis is carried out in terms of scalability, reliability, accuracy, mean average precision, throughput. The potential use of the current grid can be increased, and fault frequency can be decreased, with better monitoring technologies and predictive techniques. Proposed technique attained accuracy of 93 %, throughput of 94 %, reliability of 81 %, mean average precision of 89 %, scalability of 92 %.

Study on smart grid fault detection based on ZigBee and MapX technologies

As the complexity of the power system and the increasing demand for electricity continue to rise, ensuring the reliability and stability of the power system becomes critically important. This study utilizes ZigBee communication modules to establish a sensor network for real-time monitoring of the power system's status and employs MapX technology to achieve geographic information visualization of the power system. Through an improved RBF-PCA-WFCM algorithm, grid data can be dimensionally reduced, enabling simulation training tailored to grid loads, resulting in more accurate algorithm analysis and optimization. This paper introduces fault localization and diagnostic methods, as well as clustering optimization algorithms, to achieve fault monitoring, analysis, and identification in the smart power grid. Research findings demonstrate that the intelligent power grid data analysis and fault detection system based on ZigBee and MapX technologies can enhance the reliability of the power system and better support the increasing demands of the evolving field of electricity.

Cognitive knowledge graph generation for grid fault handling based on attention mechanism combined with multi-modal factor fusion

With the rapid development of artificial intelligence, reinforcement learning frameworks combined with knowledge graph have been widely applied in power grid fault detection and handling. However, the existing algorithms lack the consideration of multi-modal factors, which makes it difficult to improve the accuracy of prediction. To solve the above problems, we propose the cognitive knowledge graph generation algorithm for grid fault handling based on the attention mechanism combined with multi-modal factors fusion. It uses the variational auto-encoder to initialize the embedding of power equipment nodes with different modal factor constraints and dynamically fuses the embeddings of nodes based on attention mechanism, then we train the representation of nodes and relations based on TransR and construct a cognitive knowledge graph of power grid fault handling. Our algorithm is validated on the power grid fault handling data set, and the accuracy is improved.

Seamless Power Management for a Distributed DC Microgrid with Minimum Communication Links under Transmission Time Delays

To maintain voltage stabilization under transmission time delays, this paper proposes a seamless power management scheme for a distributed DC microgrid (DCMG) with minimum digital communication links (DCLs). First, a DCL topology with minimum communication data is presented for the construction of distributed DCMG system not only to mitigate the communication burden but also to enhance the system’s flexibility and reliability. In addition, based on information gathered from nearby agents and local measurements, the operating modes of local agents in a DCMG system are determined properly to ensure a proper power balance under various conditions. During normal operation, the proposed scheme works as a distributed control scheme either in the grid-connected or islanded mode to take advantage of the distributed control method. To maintain seamless power management even under transmission time delays such as grid fault detection delays and grid recovery detection delays, the operating modes of each agent in a DCMG system are switched to a decentralized scheme based on the droop control method. When the utility grid information is properly identified by all power agents after a transmission time delay, the DCMG system returns to the distributed control scheme based on DC-link voltage (DCV) control to guarantee voltage stabilization. Furthermore, the scalability issue of a distributed DCMG system is also considered in this paper when an additional energy storage system (AESS) agent is involved in the DCMG system. For this purpose, a DCL topology with minimum communication data is designed for the AESS, which enables power units to participate in or to leave the distributed DCMG system easily. Simulation and experimental results under various conditions demonstrate the effectiveness and reliability of the proposed seamless power management strategy.

Resource Orchestration of Cloud-Edge–based Smart Grid Fault Detection

Real-time smart grid monitoring is critical to enhancing resiliency and operational efficiency of power equipment. Cloud-based and edge-based fault detection systems integrating deep learning have been proposed recently to monitor the grid in real time. However, state-of-the-art cloud-based detection may require uploading a large amount of data and suffer from long network delay, while edge-based schemes do not adequately consider the detection requirement and thus cannot provide flexible and optimal performance. To solve these problems, we study a cloud-edge based hybrid smart grid fault detection system. Embedded devices are placed at the edge of the monitored equipment with several lightweight neural networks for fault detection. Considering limited communication resources, relatively low computation capabilities of edge devices, and different monitoring accuracies supported by these neural networks, we design an optimal communication and computational resource allocation method for this cloud-edge based smart grid fault detection system. Our method can maximize the processing throughput of the system and improve resource utilization while satisfying the data transmission and processing latency requirements. Extensive simulations are conducted and the results show the superiority of the proposed scheme over comparison schemes. We have also prototyped this system and verified its feasibility and performance in real-world scenarios.

Fast Fault Detection Using Optimal Intrinsic Modes and Energy Deviation Matrix in Distribution Systems

The detection, classification, and location of grid faults have always been challenging for power engineers. Among these, the detection of the grid fault is the most critical task because, without the detection, the remaining two processes will not start. In this study, a fast light-weight fault detection is proposed. This technique uses an empirical mode decomposition to decompose the fault signals into intrinsic mode functions (IMFs). Effective IMFs were selected using the earth mover’s distance-based selection algorithm. The selected IMFs are then passed through an energy calculator and then through a moving median absolute deviation stage to generate an energy deviation matrix (EDM). Based on the EDM, the occurrence of faults can be determined. The proposed technique was verified by evaluating various grid fault scenarios using the MATLAB/Simulink software platform. Furthermore, the proposed method shows a more precise and satisfactory performance than existing standard fault-detection methods.

Research on main transformer defect detection methods based on Conditional Inference Tree and AdaBoost Algorithm

With the development of data science, there are more and more ways to dig for patterns hidden in the data. If we can apply the advanced model of data science to the power data, we can mine the potential value of power data. In this paper, the monitoring and inspection data of 110kV main transformers and related basic environmental data are fused to establish conditional inference tree model and AdaBoost algorithm to evaluate whether the 110kV main transformers have defects and faults. The accuracy of the two algorithms is compared. Finally, we select the AdaBoost algorithm with higher accuracy for building the 110kV main transformer fault evaluating model. This paper provides a reference for the automatic detection of power grid faults, and provides new ideas for the application of power grid big data.

Automatic PV Grid Fault Detection System with IoT and LabVIEW as Data Logger

Fault detection of the photovoltaic (PV) grid is necessary to detect serious output power reduction to avoid PV modules’ damage. To identify the fault of the PV arrays, there is a necessity to implement an automatic system. In this IoT and LabVIEW-based automatic fault detection of 3 × 3 solar array, a PV system is proposed to control and monitor Internet connectivity remotely. Hardware component to automatically reconfigure the solar PV array from the series-parallel (SP) to the complete cross-linked array underneath partial shading conditions (PSC) is centered on the Atmega328 system to achieve maximum power. In the LabVIEW environment, an automated monitoring system is developed. The automatic monitoring system assesses the voltage drop losses present in the DC side of the PV generator and generates a decimal weighted value depending on the defective solar panels and transmits this value to the remote station through an RF modem, and provides an indicator of the faulty solar panel over the built-in Interface LabVIEW. The managing of this GUI indicator helps the monitoring system to generate a panel alert for damaged panels in the PV system. Node MCU in the receiver section enables transmission of the fault status of PV arrays via Internet connectivity. The IoT-based Blynk app is employed for visualizing the fault status of the 3 × 3 PV array. The dashboard of Blynk visualizes every array with the status.

Identification of important nodes based on dynamic evolution of inter-layer isomorphism rate in temporal networks

The identification of important nodes can not only improve the research about the structure and function of the network, but also encourage people to widely promote the application fields such as in infectious disease prevention, power grid fault detection, information dissemination control, etc. Currently, numerous conclusions have been proved on the identification of important nodes based on the static-network, which may lead the general property to be weakened as resistivity and conductivity experience the dynamic evolution of the relationship between network nodes with time. Temporal network analysis can more accurately describe the change of interaction order and interaction relationship of network nodes in the process of spatio-temporal evolution, and establish an appropriate temporal network model, as well as provide scientific theoretical support for the identification of important nodes. In this paper, we pay attention to considering the intensity of adjacent and cross-layer coupling, and propose a super-adjacency matrix (ISAM) method based on inter-layer isomorphism rate to represent the temporal networks and measure the importance of nodes. And at the same time, it is given that the temporal network <i>G</i> has <i>N</i> nodes and <i>T</i> time layers, and the ISAM is a super adjacency matrix composed of intra-layer and inter-layer relationships of adjacent and cross-layer networks, and its size is <i>NT</i> × <i>NT</i>. We focus on the study of the coupling between adjacent and cross-layer networks. The traditional method (SAM) considers the isomorphism rate of adjacent layers as a constant. In the improved method (SSAM), the connection between layers is described by a neighbor topological overlap coefficient. In this paper, the concept of the compatible similarity between cross-layer networks is given first, and then, by combining the projection value of vectors in <i>n</i>-dimensional real space and the contribution value of node neighbors, the inter-layer approximation relation coefficient of temporal network is inferred and analyzed. Generally speaking, it ensures the difference in coupling degree among different nodes in the inter-layer relationship. We calculate the importance of nodes based on eigenvector centrality in temporal network, which presents the importance of node <i>i</i> progressing with time. Simultaneously, the robustness of temporal network is studied by making use of the difference in temporal global efficiency. In the end, the operator of Kendall correlation coefficient is used to evaluate the node ranking effect of different time layers between the eigenvector-based centrality and the difference of temporal global efficiency. According to the experimental results of ISAM, SSAM and SAM on Workspace and Email-eu-core data sets, the average Kendall <i>τ</i> of both ISAM methods considering adjacent and cross-layer network isomorphism rate can be increased by 8.37% and 2.99% respectively. The conclusions show that the measurement method of temporal network inter-layer isomorphism rate is reliable and effective.

Autonomous Model Predictive Controlled Smart Inverter With Proactive Grid Fault Ride-Through Capability

This article presents an autonomous model predictive controlled smart photovoltaic (PV) inverter with proactive grid fault-ride through capability. The proposed smart inverter control features decoupled active and reactive power. It can seamlessly switch between the anticipated modes of operation based on grid condition or grid operator command. The smart inverter autonomously adjusts its active and reactive power set-points according to the grid condition, it operates in maximum power point tracking (MPPT) and low voltage ride through (LVRT) modes in normal grid and faulty grid condition modes, respectively. The proposed novel autonomous model predictive control (AMPC) scheme is leveraged to enhance the operation of the smart inverter. The AMPC includes online weight factor auto-tuning and control objective normalization to eliminate the required trial-and-error weight factor design stage in conventional model predictive control. This feature is particularly beneficial to the dual-mode smart inverter operation. The performance of the proposed grid-tied smart inverter based on the AMPC is verified experimentally. The results demonstrate that the proposed AMPC-based smart inverter features robust grid fault detection, autonomous adjustment of active and reactive power set-points, seamless transition between modes of operation, and elimination of controller tuning effort.

A New Protection Scheme for an SSSC in an MV Network by Using a Varistor and Thyristors

To control power flow and manage fault level in meshed MV networks, back-to-back voltage source converters (B2B-VSCs) are being used. However, their high cost and relatively low efficiency are of concerns. Partially rated series compensators, such as SSSCs or UPFCs, are desired but come with the challenge of protecting the device during grid faults. Their potential of use has been limited in comparison with the fully rated back-to-back converters. This paper proposes a new system topology including thyristor crowbars and a varistor to protect the SSSC in an MV network and improve the reliability and flexibility of the network operation. Using the proposed method, the time required for isolating the series compensator from the grid is reduced from at least 20 ms, corresponding to the interruption time of conventional circuit breakers, down to 3 μs in the worst case in addition to the grid fault detection delay. The performance is evaluated by simulation. A small-scale single-phase prototype operating at 230 V/16 A is tested in order to demonstrate the concept.

Multi-stage fault warning for large electric grids using anomaly detection and machine learning

In the monitoring of a complex electric grid, it is of paramount importance to provide operators with early warnings of anomalies detected on the network, along with a precise classification and diagnosis of the specific fault type. In this paper, we propose a novel multi-stage early warning system prototype for electric grid fault detection, classification, subgroup discovery, and visualization. In the first stage, a computationally efficient anomaly detection method based on quartiles detects the presence of a fault in real time. In the second stage, the fault is classified into one of nine pre-defined disaster scenarios. The time series data are first mapped to highly discriminative features by applying dimensionality reduction based on temporal autocorrelation. The features are then mapped through one of three classification techniques: support vector machine, random forest, and artificial neural network. Finally in the third stage, intra-class clustering based on dynamic time warping is used to characterize the fault with further granularity. Results on the Bonneville Power Administration electric grid data show that i) the proposed anomaly detector is both fast and accurate; ii) dimensionality reduction leads to dramatic improvement in classification accuracy and speed; iii) the random forest method offers the most accurate, consistent, and robust fault classification; and iv) time series within a given class naturally separate into five distinct clusters which correspond closely to the geographical distribution of electric grid buses.

Big data outlier detection model based on improved density peak algorithm

Outlier detection is an important branch of data mining. This paper proposes an advanced fast density peak outlier detection algorithm based on the characteristics of big data. The algorithm is an outlier detection method based on the improved density peak clustering algorithm. This paper improves the original algorithm. From the perspective of outlier detection, although it is a clustering idea, it avoids the clustering process, reduces the time complexity of the cluster-based outlier detection algorithm, and absorbs. The outlier detection based on neighbors is not sensitive to data dimensions and other advantages. In the power industry, outlier detection can be used in areas such as grid fault detection, equipment fault detection, and power abnormality detection. The simulation experiment of outlier detection based on the daily load curve of single and multiple transformers in a certain province shows that the improved algorithm can effectively detect outliers in the data.

Fault Detection in Grid Connected Wind Energy Conversion Systems

This paper develops an artificial neural network-based implementation for detecting fault in grid connected Wind energy conversion system. The proposed algorithm that would predict the fault that occurs on the grid connected system is completely automated using the ANN algorithm. The fault in the grid is considered to implement the proposed algorithm for identify the fault. The automation is carried out using Back Propagation Network Algorithm (BPNA) and MATLAB based realization using Simulink and M-file functions is carried out and the results are tabulated. The efficient training algorithm and the testing is carried out on the grid connected WECS. The parameters accuracy of this algorithm is analyzed with previous implementations. The outcome of the proposed implementation provided satisfactory results.

An Improved Power Management Strategy for MAS-Based Distributed Control of DC Microgrid under Communication Network Problems

In this paper, an improved power management strategy (PMS) for multi-agent system (MAS)-based distributed control of DC microgrid (DCMG) under communication network problems is presented in order to enhance the reliability of DCMG and to ensure the system power balance under various conditions. To implement MAS-based distributed control, a communication network is constructed to exchange information among agents. Based on the information obtained from communication and local measurements, the decision for the local controller and communication is optimally given to guarantee the system power balance under various conditions. The operating modes of the agents can be determined locally without introducing any central controller. Simultaneously, the agents can operate in a deliberative and cooperative manner to ensure global optimization by means of the communication network. Furthermore, to prevent the system power imbalance caused by the delay in grid fault detection and communication in case of the grid fault, a DC-link voltage (DCV) restoration algorithm is proposed in this study. In addition, to avoid the conflict in the DCV control among power agents in case of the grid recovery under communication failure, a grid recovery identification algorithm is also proposed to improve the reliability of DCMG operation. In this scheme, a special current pattern is generated on the DC-link at the instant of the grid recovery by the grid agent, and other power agents identify the grid recovery by detecting this current pattern. Comprehensive simulations and experiments based on DCMG testbed have been carried out to prove the effectiveness of the PMS and the proposed control schemes under various conditions.

Analysis of Local Measurement-Based Algorithms for Fault Detection in a Multi-Terminal HVDC Grid

One of the most important challenges of developing multi-terminal (MT) high voltage direct current (HVDC) grids is the system performance under fault conditions. It must be highlighted that the operating time of the protection system needs to be shorter than a few milliseconds. Due to this restrictive requirement of speed, local measurement based algorithms are mostly used as primary protection since they present an appropriate operation speed. This paper focuses on the analysis of local measurement based algorithms, specifically overcurrent, undervoltage, rate-of-change-of-current, and rate-of-change-of-voltage algorithms. A review of these fault detection algorithms is presented. Furthermore, these algorithms are applied to a multi-terminal grid, where the influence of fault location and fault resistance is assessed. Then, their performances are compared in terms of detection speed and maximum current interrupted by the HVDC circuit breakers. This analysis aims to enhance the protection systems by facilitating the selection of the most suitable algorithm for primary or backup protection systems. In addition, two new fault type identification algorithms based on the rate-of-change-of-voltage and rate-of-change-of-current are proposed and analyzed. The paper finally includes a comparison between the previously reviewed local measurement based algorithms found in the literature and the simulation results of the present work.