Fault Diagnosis In Distribution Networks Research Articles

Active Distribution Network Fault Diagnosis Based on Improved Northern Goshawk Search Algorithm

Timely and accurate fault location in active distribution networks is of vital importance to ensure the reliability of power grid operation. However, existing intelligent algorithms applied in fault location of active distribution networks possess slow convergence speed and low accuracy, hindering the construction of new power systems. In this paper, a new regional fault localization method based on an improved northern goshawk search algorithm is proposed. The population quality of the samples was improved by using the chaotic initialization strategy. Meanwhile, the positive cosine strategy and adaptive Gaussian–Cauchy hybrid variational perturbation strategy were introduced to the northern goshawk search algorithm, which adopted the perturbation operation to interfere with the individuals to increase the diversity of the population, contributing to jumping out of the local optimum to strengthen the ability of local escape. Finally, simulation verification was carried out in a multi-branch distribution network containing distributed power sources. Compared with the traditional regional localization models, the new method proposed possesses faster convergence speed and higher location accuracy under different fault locations and different distortion points.

Detection of single-phase-to-ground faults in distribution networks based on Gramian Angular Field and Improved Convolutional Neural Networks

Fault detection of transmission lines is a fundamental guarantee for the stable operation of the power grid. The characteristic information is not obvious when a single-phase-to-ground fault occurs in the complex distribution network, and the existing methods are vulnerable to the problems of fault conditions and environmental noise. To address this problem, a new method of distribution network fault diagnosis is proposed based on Gramian Angular Field (GAF) and Improved Convolutional Neural Networks (ICNN) in this paper. Firstly, the time-series signals are normalized and transformed to a polar coordinate system. The encoded signals are converted into visual virtual images through the inner product calculation of GAF algorithm. Secondly, multi-scale module and attention mechanism are added to the convolutional layer of the ICNN. Among them, the multi-scale module, which enhances the local feature extraction capability for the original image, is composed of multiple convolutional channels. The purpose of attention mechanism is to help the weight distribution and depth feature filtering of ICNN, focusing on the key features between fault data. Finally, the transfer learning method is further introduced to improve the fault diagnosis model performance for different online situations. The experimental results show that the overall accuracy of the proposed fault location method is close to 99% under different topological situations. Even with the influence of ambient noise, the accuracy of GAF-ICNN still approaches 97.5%.

Distribution network fault diagnosis method for the deep integration of cyber-physics

• A distribution network fault location method for deep integration of cyber-physics is proposed. • The complexity of fault diagnosis is reduced by building a library of faulty suspect components. • The characteristic data collected by the D-PMUs is used and improves the speed of fault diagnosis. • Petri net model is used to accurately describe the process of fault occurrence and improves the accuracy. The deep integration of cyber-physics provides multi-source information data for the fault analysis of distribution networks, but with conventional methods, it is difficult to select and use these data reasonably. This paper proposes a fault diagnosis method for a distribution network based on a distributed phasor measurement unit (D-PMU) and a Petri net. The method introduces a fault suspect component library formed with coupling pre-judgment and a fault diagnosis model, which simplifies the process of network analysis and improves the speed of the fault diagnosis. The method makes full use of the time constraint relationship of the alarm information, correctly selects the action time point, and achieves an improvement of the fault diagnosis accuracy. In addition, considering the problem that the measurement equipment coverage is not comprehensive in the actual distribution network, the proposed method inputs the characteristic information collected by the D-PMU at the typical measurement point into the Petri net model, analyzes the fault of the distribution network, and avoids the problem of the scheduling being risky due to the result of the probabilistic form. Finally, this paper describes the setup of some experimental analyses to verify the effectiveness of the proposed method.

Modeling and fault diagnosis of distribution networks cyber physical system based on IEC61850

In order to improve the fault tolerance and solution efficiency of power grid fault diagnosis model, a distribution network cyber physical system (CPS) fault diagnosis model based on advanced digital power system simulator (ADPSS) and IEC61850 is proposed. Firstly, the action rules and action status of CPS fault mode of distribution network from the whole correlation of fault element, protection action and circuit breaker trip, and the CPS grid topology of distribution network is constructed in ADPSS based on the topology and component characteristics of system. Secondly, according to the IEC61850 unified information model and modeling rules for the protection function of system equipment and operation function of circuit breaker, an information model with alarm, measurement, protection and fault recording logic is established. Then, the action status is obtained and used for the establishment of the distribution network fault diagnosis model by parsing the GOOSE message. Finally, the proposed modeling method and fault diagnosis strategy is validated through the test case.

Deep learning technology of Internet of Things Blockchain in distribution network faults

AbstractNowadays, the development of human society and daily life are inseparable from the power supply. Therefore, people also put forward higher requirements for the reliability of distribution network, but power companies can only passively deal with distribution network failures, which is a bottleneck for the improvement of distribution network reliability. The Internet of Things (IoT) is the best solution for online equipment status monitoring and basic data sharing for large, widely distributed, relatively fixed, and large numbers of equipment. The construction of the IoT for power distribution equipment faces many important problems, including the selection of networking, equipment selection, and interaction standards. When researching the implementation plan, research on the distribution of IoT market was carried out. Based on the grid, the idea of optimizing the investment selection plan of the power distribution using IoT was discussed, and a result verification model was established. After the completion of the theoretical part, a case study of medium-voltage distribution grid equipment management and medium-voltage distribution network equipment management based on the grid was carried out by taking a real enterprise application situation as an example. Realizing fault diagnosis of distribution network will not only provide decision support for operation and maintenance of distribution network for power companies, but also have great economic and social benefits. Aiming at the shortcomings of single data mining method in distribution network fault diagnosis, hybrid data mining method is proposed. First, rough set theory is used to reduce the original fault data and form a simplified rule set. Because of the non-linearity of distribution network fault and the strong learning ability, adaptability, and robustness of Bayesian network, Bayesian network can be used to classify distribution network faults. Therefore, a simplified fault diagnosis system is established in this paper, and its correctness is confirmed. Then, the learning and training are carried out by using Bayesian network to call the simplest rule set, which has the characteristics of short learning and training time and high diagnostic accuracy.

Current Sag and Mismatch Based Earth Fault Location for Distribution Network with Renewable Energy Resources

Fault diagnosis is increasingly becoming more challenging in distribution network due to complicated network topology and coupling disturbances introduced into the systems. This requires a higher level of rapidity and accuracy of fault diagnosis, location and recovery. This paper proposes an earth fault location algorithm for distribution network using current analysis. The fault information can be obtained by those measuring equipment that have already installed in critical nodes and generations. Then, the mechanism model is proposed to illustrate the equilibrium relationship of current flow in distribution network with renewable energy sources. The criterions named current sag and current mismatch are also proposed in this paper for fast fault area location and accurate fault node location. Assessment of this method is carried out using a distribution network model with renewable energy sources in PSCAD. The experimental results confirm the effectiveness and superiority of the proposed method in distribution network fault diagnosis over conventional methods.

Identification of distribution network topology parameters based on multidimensional operation data

The connection relationship of distribution network topology is of great significance for the maintenance and fault diagnosis of distribution network, and scheduled power outage optimization. At present, the verification of topological documents mainly relies on on-site inspection, which consumes a lot of manpower and material resources and is inefficient. Therefore, an efficient method for topology verification of low-voltage substation areas is required. Given this background, a model for error correction and user access phase identification of low-voltage stations based on multi-dimensional voltage data collected by smart meters is presented in this paper, which can provide a certain reference for topology identification and line troubleshooting of low-voltage substations. First, the t-Distributed Stochastic Neighbor Embedding (t-SNE) algorithm and the Principal Component Analysis (PCA) performs dimensionality reduction on the original load data to solve the problem of redundancy caused by the high dimension of the original voltage data set. Second, the Local Outlier Factor (LOF) algorithm is used to identify abnormal samples in the voltage data set. Then, the spectral clustering method is used to cluster the dimensionality-reduced load data to realize the phase identification of single-phase users in the low-voltage station area. Finally, the real data of a certain area in Haining, Zhejiang Province of China are used as simulation cases for demonstrating. The results of the case studies show that the model proposed in this paper is feasible and effective.

An edge computing method using a novel mode component for power transmission line fault diagnosis in distribution network

The commonly used fault diagnosis method of the power transmission line (PTL) is the traveling wave method which uses the traditional mode components $$\alpha $$ , $$\beta $$ and o to realize the fault type recognizing and the fault location in distribution network. However, the traditional mode components of the traveling wave method may take the value of zero that cannot provide the effective fault feature for the A-phase ground fault, or the A and B double phases short-circuit fault. Moreover, a single traditional mode component cannot recognize all the types of the faults, and the fault recognition has to use the multiple mode components that increases the complexity. To address the aforementioned issues, this paper proposes an edge computing scheme based on a novel mode component $$\gamma $$ for PTL fault diagnosis in distribution network. Firstly, the novel mode component $$\gamma $$ is fused according to the traditional mode components $$\alpha $$ and $$\beta $$ , then the propagation property of the traveling wave- $$\gamma $$ is derived. Secondly, the edge computing scheme and the PTL fault diagnosis method are designed for distribution network by using the only novel mode component $$\gamma $$ . Thirdly, the relationship table of the fault type and the boundary conditions as well as the component $$\gamma $$ expressions is derived for the fault recognition using the only traveling wave- $$\gamma $$ . In addition, the fault distance computing is explored by using the traveling wave- $$\gamma $$ . The computing is composed with the wave- $$\gamma $$ head identification using the optimal Wavelet vanishing moment parameter, and the arrival time determination, and the fault location method. To evaluate the efficiency of the proposed edge computing scheme, the simulation experiment and the laboratory tests are conducted. The simulation experiment results show that, the novel mode component $$\gamma $$ not only provides the effective fault features for all the ten types of the power transmission line faults, but also increases fault locating accuracy by the average 2.08 % compared with the methods using the traditional mode component $$\alpha $$ , $$\beta $$ and o. Meanwhile, the laboratory tests show that the edge computing method is practical.

Two-Terminal Fault Location Method of Distribution Network Based on Adaptive Convolution Neural Network

When a single-phase ground fault occurs in a distribution network, it is generally allowed to operate with faults for one to two hours, which may lead to further development of the fault and even threaten the safe operation of the power system. Therefore, when a small current system has a ground fault, it must be quickly diagnosed to shorten the time of operation with fault. In this paper, an adaptive convolutional neural network (ACNN)-based fault line selection method is proposed for a distribution network. This method improves the feature extraction ability of the network by improving the pooling model. Compared with deep belief network (DBN), it can improve the accuracy of fault classification by 7.86% and reduce the training time by 42.7%. On this basis, the secondary fault location is identified using the principle of two-terminal fault location. In this research, fault data obtained by Simulink simulation is used as training set, and ACNN model is built based on TensorFlow framework. The analysis of results proves that the model has a high fault recognition rate and fast convergence speed. It can be used as an auxiliary hand for fault diagnosis in distribution networks.

A Minimum Hitting Set Algorithm With Prejudging Mechanism for Model-Based Fault Diagnosis in Distribution Networks

Fault diagnosis has a critical role in maintaining the reliability and stability of distribution networks. Compared with common diagnosis methods, such as expert systems that depend on the professional experience and information of fault, the model-based diagnosis (MBD) has many distinct advantages. However, the efficiency and accuracy of the calculation of the minimum hitting set (MHS) limit the performance of MBD. To overcome these limitations, an MHS algorithm with a prejudging mechanism is proposed in this article. This mechanism is constructed based on the distribution characteristics of MHS. The use of this mechanism prior to MHS calculation enables the removal of most non-HS particles and all super-MHS particles and enables the compression of the MHS calculation solution space several times. Moreover, after compression, only two types of particles remain in the left solution space, which are non-HS and MHS. As a result, the MHS can be picked out directly through the HS fitness function and the HS criterion and by avoiding the participation of the MHS ensured strategy (MHSES). Three cases studies have been introduced to test the effect of the prejudging mechanism on the performance of the MHS calculation. In addition, a discussion on the MBD of the distribution networks has been presented to validate the efficiency and accuracy of the proposed algorithm.

A New Search Algorithm of MBD Based on Spider Web and Its Application in Power Distribution Network Fault Diagnosis

To reduce the spatial complexity and search time of minimum hitting sets in model-based diagnosis (MBD), a new algorithm for searching minimum hitting sets of MBD is proposed in this paper, which can use the characteristics of minimum hitting sets and the idea of spider prey in biology. We call it cobweb search algorithm. In the algorithm, the generation of visit spiders and search strategy are proposed, and the visit spider that can find the visit path of cobweb to search all minimum hitting sets within a cobweb is constructed. Based on the experiment comparisons, Cobweb search algorithm has better performance than other algorithms of searching minimum hitting sets. As an example, a 14-node power distribution network model is constructed. The diagnosis process with MBD is introduced and analyzed in detail, at the same time the cobweb search algorithm is applied in power distribution network fault diagnosis. The experiment results verify the effectiveness and superiority of the proposed algorithm.

A parallel heuristic reduction based approach for distribution network fault diagnosis

Abstract For the large volume of data in power systems, existing approaches to rough sets reduction either run on a single machine, or are paralleledly achieved in an approximate manner. They also seldom consider rough sets based value reduction. These problems restrict them in applications of power systems. In order to accelerate attribute reduction and value reduction, and improve the efficiency of fault diagnosis analysis in power systems, in this paper, we present a parallel heuristic approach to exact attribute reduction and value reduction for fault diagnosis in distribution network. We obtain the diagnosis rules and diagnose and locate the faults in the distribution network. Our parallel algorithms have been implemented on the MapReduce platform. The experimental results show that our method can effectively improve reduction process and improve the accuracy of reduction results in dealing with a large volume of data sets.

Researching on Distribution Network Fault Location System

In this study, we propose the method using rough set theory to realize the distribution network fault positioning in WEBGIS environment. According to the distribution network tree structure, using of the area user s’ fault complain information as condition attributes, fault elements as decision attribute, form the decision table automatically. By using rough set method reduce the decision table, the minimal reduction of decision table for fault diagnosis is derived, the minimal diagnostic rules are obtained, guarantee the objectivity of the rule. When the fault complain call information is imperfect, it can still achieve rapid, accurate fault location on purpose, has good fault tolerance performance. In this study the use of C# language programming to achieve the reduction process for distribution network fault diagnosis decision table, combined with the WEBGIS platform, make full use of internal network database resources, to realize fault diagnosis simple and rapid operation and fault location visualization. The results show that the method is feasible and effective.

Fault diagnosis in distribution networks with distributed generation

The penetration of distributed generation (DG) in distribution power system would affect the traditional fault current level and characteristics. Consequently, the traditional protection arrangements developed in distribution utilities are difficult in coordination. Also, the reclosing scheme would be affected. With the rapid developments in distribution system automation and communication technology, the protection coordination and reclosing scheme based on information exchange for distribution power system can be realized flexibly. This paper proposes a multi-agent based scheme for fault diagnosis in power distribution networks with distributed generators. The relay agents are located such that the distribution network is divided into several sections. The relay agents measure the bus currents at which they are located such that it can detect and classify the fault, and determine the fault location. The proposed technique uses the entropy of wavelet coefficients of the measured bus currents. The performance of the proposed protection scheme is tested through simulation of two systems. The first system is a benchmark medium voltage (MV) distribution system and the second system is practical 66 kV system of the city of Alexandria.

Field studies using a neural-net-based approach for fault diagnosis in distribution networks

The paper discusses results of studies performed on a new fault-diagnosis method for distribution systems using acquired field data. The effectiveness of the fault-diagnosis method in distinguishing between faulted conditions and system conditions that appear fault-like is demonstrated, for a field-test system, using data recorded at two utility distribution systems. The new method uses two major components: a signal preprocessor and a novel supervised clustering-based neural network which perform fault detection in the presence of arcing, classification of the fault type and preliminary fault location through the identification of the faulted phase. The work represents the first time that a supervised clustering neural network has been used for distribution fault diagnosis.

95/03676 Machine learning and knowledge building for fault diagnosis in distribution network

95/03676 Machine learning and knowledge building for fault diagnosis in distribution network

Machine learning and knowledge building for fault diagnosis in distribution network

A fault diagnosis system for distribution networks using machine learning of fault patterns is described. The learning process can be implemented during fault condition on the actual network or through a distribution network simulator. Knowledge acquisition is almost automated. The proposed system can be integrated into the existing SCADA system to provide a guide for engineers during fault conditions.

Using fuzzy sets to model the uncertainty in the fault location process of distribution networks

In the computerized fault diagnosis of distribution networks the heuristic knowledge of the control center operators can be combined with the information obtained from the network database and SCADA system. However, the nature of the heuristic knowledge is inexact and uncertain. Also the information obtained from the remote control system contains uncertainty and may be incorrect, conflicting or inadequate. This paper proposes a method based on fuzzy set theory to deal with the uncertainty involved in the process of locating faults in distribution networks. The method is implemented in a prototype version of the distribution network operation support system. >

The interface between a conventional database and an expert system application

Efficient fault locating and network reconfigurating in medium-voltage distribution networks call for flexible connection with the SCADA (system central and data acquisition) and network database systems. Using the information from these systems together with the heuristic knowledge of the control centre operator a high-level knowledge-based assistance system can be built for fault diagnosis of distribution networks. Such a system is being developed by the Power Engineering Group at Tampere University of Technology. This paper gives a detailed description of the dynamic frame-based model of the distribution network and outlines the use of the interface between the network database system and the expert system for fault diagnosis in a fault situation.