Arc Grounding Fault Identification Using Integrated Characteristics in the Power Grid

Arc grounding faults occur frequently in the power grid with small resistance grounding neutral points.The existing arc fault identification technology only uses the fault line signal characteristics to set the identification index,which leads to detection failure when the arc zero-off characteristic is short.To solve this problem,this paper presents an arc fault identification method by utilizing integrated signal characteristics of both the fault line and sound lines.Firstly,the waveform characteristics of the fault line and sound lines under an arc grounding fault are studied.After that,the convex hull,gradient product,and correlation coefficient index are used as the basic characteristic parameters to establish fault identification criteria.Then,the logistic regression algorithm is employed to deal with the reference samples,establish the machine discrimination model,and realize the discrimination of fault types.Finally,simulation test results and experimental results verify the accuracy of the proposed method.The comparison analysis shows that the proposed method has higher recognition accuracy,especially when the arc dissipation power is smaller than 2×10^(3) W,the zero-off period is not obvious.In conclusion,the proposed method expands the arc fault identification theory.

Analysis and Identification on Fault of Rub-Impact between Rotor and Stator

According to the background of the rub impact faults of aerial engines and industrial turbines, two kinds of test rigs, on the base of the dynamics model, are established to study the rub impact faults between rotor and stator with free supports. The orbit of the vibration of rotor displacement is respectively examined on the four impact conditions, which are the normal state with no impact, the early sharp impact statement, the semi sharp impact statement and the terminal blunt impact statement. The route to chaos, appearing with the early sharp impact, is observed for the first time. By analyzing the frequency domain characteristics of the experimental data on four impact conditions, it is testified that the appearance of the sub harmonic vibrations of the order 1/3 and 1/4 is the effective evidence to judge whether or not the blade has initial light rub impact. When there are only the harmonic vibrations of the order of 1/1 and 1/2, the blade stator rub impact faults have become very serious.

Identification of abnormal conditions in high-dimensional chemical process based on feature selection and deep learning

Identification of abnormal conditions is essential in the chemical process.With the rapid development of artificial intelligence technology,deep learning has attracted a lot of attention as a promising fault identification method in chemical process recently.In the high-dimensional data identification using deep neural networks,problems such as insufficient data and missing data,measurement noise,redundant variables,and high coupling of data are often encountered.To tackle these problems,a feature based deep belief networks(DBN)method is proposed in this paper.First,a generative adversarial network(GAN)is used to reconstruct the random and non-random missing data of chemical process.Second,the feature variables are selected by Spearman’s rank correlation coefficient(SRCC)from high-dimensional data to eliminate the noise and redundant variables and,as a consequence,compress data dimension of chemical process.Finally,the feature filtered data is deeply abstracted,learned and tuned by DBN for multi-case fault identification.The application in the Tennessee Eastman(TE)process demonstrates the fast convergence and high accuracy of this proposal in identifying abnormal conditions for chemical process,compared with the traditional fault identification algorithms.

Fault Location Identification for Localized Intermittent Connection Problems on CAN Networks

The intermittent connection（IC）of the field-bus in networked manufacturing systems is a common but hard troubleshooting network problem,which may result in system level failures or safety issues.However,there is no online IC location identification method available to detect and locate the position of the problem.To tackle this problem,a novel model based online fault location identification method for localized IC problem is proposed.First,the error event patterns are identified and classified according to different node sources in each error frame.Then generalized zero inflated Poisson process（GZIP）model for each node is established by using time stamped error event sequence.Finally,the location of the IC fault is determined by testing whether the parameters of the fitted stochastic model is statistically significant or not using the confident intervals of the estimated parameters.To illustrate the proposed method,case studies are conducted on a 3-node controller area network（CAN）test-bed,in which IC induced faults are imposed on a network drop cable using computer controlled on-off switches.The experimental results show the parameters of the GZIP model for the problematic node are statistically significant（larger than 0）,and the patterns of the confident intervals of the estimated parameters are directly linked to the problematic node,which agrees with the experimental setup.The proposed online IC location identification method can successfully identify the location of the drop cable on which IC faults occurs on the CAN network.

A Successive Shift Box-Counting Method for Calculating Fractal Dimensions and Its Application in Identification of Faults

Fractal dimensions of a terrain quantitatively describe the self-organizedstructure of the terrain geometry. However, the local topographic variation cannot be illustrated bythe conventional box-counting method. This paper proposes a successive shift box-counting method,in which the studied object is divided into small sub-objects that are composed of a series of gridsaccording to its characteristic scaling. The terrain fractal dimensions in the grids are calculatedwith the successive shift box-counting method and the scattered points with values of fractaldimensions are obtained. The present research shows that the planar variation of fractal dimensionsis well consistent with fault traces and geological boundaries.

An Efficient Fuzzy Logic Fault Detection and Identification Method of Photovoltaic Inverters

Fuzzy logic control(FLC)systems have found wide utilization in several industrial applications.This paper proposes a fuzzy logic-based fault detection and identification method for open-circuit switch fault in grid-tied photovoltaic(PV)inverters.Large installations and ambitious plans have been recently achieved for PV systems as clean and renewable power generation sources due to their improved environmental impacts and availability everywhere.Power converters represent the main parts for the grid integration of PV systems.However,PV power converters contain several power switches that construct their circuits.The power switches in PV systems are highly subjected to high stresses due to the continuously varying operating conditions.Moreover,the grid-tied systems represent nonlinear systems and the system model parameters are changing continuously.Consequently,the grid-tied PV systems have a nonlinear factor and the fault detection and identification(FDI)methods based on using mathematical models become more complex.The proposed fuzzy logic-based FDI(FL-FDI)method is based on employing the fuzzy logic concept for detecting and identifying the location of various switch faults.The proposed FL-FDI method is designed and extracted from the analysis and comparison of the various measured voltage/current components for the control purposes.Therefore,the proposed FL-FDI method does not require additional components or measurement circuits.Additionally,the proposed method can detect the faulty condition and also identify the location of the faulty switch for replacement and maintenance purposes.The proposed method can detect the faulty condition within only a single fundamental line period without the need for additional sensors and/or performing complex calculations or precise models.The proposed FL-FDI method is tested on the widely used T-type PV inverter system,wherein there are twelve different switches and the FDI process represents a challenging task.The results shows the superior and accurate performance of the proposed FL-FDI method.

Fault Identification and Health Monitoring of Gas Turbine Engines Using Hybrid Machine Learning-based Strategies

Ahealth monitoring scheme is developed in this work by using hybrid machine learning strategies to iden-tify the fault severity and assess the health status of the aircraft gas turbine engine that is subject to component degrada-tions that are caused by fouling and erosion.The proposed hybrid framework involves integrating both supervised recur-rent neural networks and unsupervised self-organizing maps methodologies,where the former is developed to extract ef-fective features that can be associated with the engine health condition and the latter is constructed for fault severity modeling and tracking of each considered degradation mode.Advantages of our proposed methodology are that it ac-complishes fault identification and health monitoring objectives by only discovering inherent health information that are available in the system I/O data at each operating point.The effectiveness of our approach is validated and justified with engine data under various degradation modes in compressors and turbines.

FAULT IDENTIFICATION IN HETEROGENEOUS NETWORKS USING TIME SERIES ANALYSIS

Fault management is crucial to pro vi de quality of service grantees for the future networks, and fault identification is an essential part of it. A novel fault identification algorithm is proposed in this paper, which focuses on the anomaly detection of network traffic. Since the fault identification has been achieved using statistical information in mana gement information base, the algorithm is compatible with the existing simple ne twork management protocol framework. The network traffic time series is verified to be non-stationary. By fitting the adaptive autoregressive model, the series is transformed into a multidimensional vector. The training samples and identif iers are acquired from the network simulation. A k-nearest neighbor classif ier identifies the system faults after being trained. The experiment results are consistent with the given fault scenarios, which prove the accuracy of the algo rithm. The identification errors are discussed to illustrate that the novel faul t identification algorithm is adaptive in the fault scenarios with network traff ic change.

Intelligent identification method and application of seismic faults based on a balanced classification network

This study combined fault identification with a deep learning algorithm and applied a convolutional neural network(CNN)design based on an improved balanced crossentropy(BCE)loss function to address the low accuracy in the intelligent identification of seismic faults and the slow training speed of convolutional neural networks caused by unbalanced training sample sets.The network structure and optimal hyperparameters were determined by extracting feature maps layer by layer and by analyzing the results of seismic feature extraction.The BCE loss function was used to add the parameter which is the ratio of nonfaults to the total sample sets,thereby changing the loss function to find the reference of the minimum weight parameter and adjusting the ratio of fault to nonfault data.The method overcame the unbalanced number of sample sets and improved the iteration speed.After a brief training,the accuracy could reach more than 95%,and gradient descent was evident.The proposed method was applied to fault identification in an oilfield area.The trained model can predict faults clearly,and the prediction results are basically consistent with an actual case,verifying the effectiveness and adaptability of the method.

Single Phase-to-Ground Fault Line Identification and Section Location Method for Non-Effectively Grounded Distribution Systems Based on Signal Injection

A diagnostic signal current trace detecting based single phase-to-ground fault line identifica- tion and section location method for non-effectively grounded distribution systems is presented in this paper.A special diagnostic signal current is injected into the fault distribution system,and then it is detected at the outlet terminals to identify the fault line and at the sectionalizing or branching point along the fault line to locate the fault section.The method has been put into application in actual distribution network and field experience shows that it can identify the fault line and locate the fault section correctly and effectively.

Misfire identification of automobile engines based on wavelet packet and extreme learning machine

Due to non-stationary characteristics of the vibration signal acquired from cylinder head,a misfire fault diagnosis system of automobile engines based on correlation coefficient gained by wavelet packet and extreme learning machine(ELM)is proposed.Firstly,the original signal is decomposed by wavelet packet,and correlation coefficients between the reconstructed signal of each sub-band and the original signal as well as the energy entropy of each sample are obtained.Then,the eigenvectors established by the correlation coefficients method and the energy entropy method fused with kurtosis are inputted to the four kinds of classifiers including BP neural network,KNN classifier,support vector machine and ELM respectively for training and testing.Experimental results show that the method proposed in this paper can effectively reflect the differences that the fault produces and identify the single-cylinder misfire accurately,which has the advantages of higher accuracy and shorter training time.

Identification of Type of a Fault in Distribution System Using Shallow Neural Network with Distributed Generation

A distributed generation system(DG)has several benefits over a traditional centralized power system.However,the protection area in the case of the distributed generator requires special attention as it encounters stability loss,failure re-closure,fluctuations in voltage,etc.And thereby,it demands immediate attention in identifying the location&type of a fault without delay especially when occurred in a small,distributed generation system,as it would adversely affect the overall system and its operation.In the past,several methods were proposed for classification and localisation of a fault in a distributed generation system.Many of those methods were accurate in identifying location,but the accuracy in identifying the type of fault was not up to the acceptable mark.The proposed work here uses a shallow artificial neural network(sANN)model for identifying a particular type of fault that could happen in a specific distribution network when used in conjunction with distributed generators.Firstly,a distribution network consisting of two similar distributed generators(DG1 and DG2),one grid,and a 100 Km distribution line is modeled.Thereafter,different voltages and currents corresponding to various faults(line to line,line to ground)at different locations are tabulated,resulting in a matrix of 500×18 inputs.Secondly,the sANN is formulated for identifying the types of faults in the system in which the above-obtained data is used to train,validate,and test the neural network.The overall result shows an unprecedented almost zero percent error in identifying the type of the faults.

A Novel Parsimonious Neurofuzzy Model Applied to Railway Carriage System Identification and Fault Diagnosis

In this paper, we suggest a novel parsimonious neurofuzzy model realized by RBFNs for railway carriage system identification and fault diagnosis. To overcome the curse of dimensionality resulting from high dimensional input variables, in our developed model the features extracted from the available observations are regarded as the input variables by adopting the higher-order statistics(HOS) technique. Such a constructed model is also applied to a practical railway carriage system, simulation results indicate that the developed neurofuzzy model possesses strong identification and fault diagnosis ability.

Fault Identification and Islanding in DC Grid Connected PV System

Nowadays, the DC distribution system has been suggested, as a replacement for the AC power distribution system with electric propulsion. This idea signifies a fresh approach of issuing energy for low-voltage installations. It can be used for any electrical application up to 20 MW and works at a nominal voltage of 1000 V DC. The DC distribution system is just an extension of the multiple DC links that previously available in all propulsion and thruster drives, which typically comprise more than 80% of the electrical power consumption on electric propulsion vessels. A fault detection and islanding scheme for DC grid connected PV system is presented in this paper. Unlike traditional ac distribution systems, protection has been challenging for dc systems. The goals of this paper are to classify and detect the fault in the PV system as well as DC grid and to isolate the faulted section so that the system keeps operating without disabling the entire system. The results show the measured values of power at PV panel and DC grid side under different fault condition, which indicates the type of fault that occurs in the system.

Combinatorial reasoning-based abnormal sensor recognition method for subsea production control system

The subsea production system is a vital equipment for offshore oil and gas production.The control system is one of the most important parts of it.Collecting and processing the signals of subsea sensors is the only way to judge whether the subsea production control system is normal.However,subsea sensors degrade rapidly due to harsh working environments and long service time.This leads to frequent false alarm incidents.A combinatorial reasoning-based abnormal sensor recognition method for subsea production control system is proposed.A combinatorial algorithm is proposed to group sensors.The long short-term memory network(LSTM)is used to establish a single inference model.A counting-based judging method is proposed to identify abnormal sensors.Field data from an offshore platform in the South China Sea is used to demonstrate the effect of the proposed method.The results show that the proposed method can identify the abnormal sensors effectively.

Review of Artificial Intelligence for Oil and Gas Exploration: Convolutional Neural Network Approaches and the U-Net 3D Model

Deep learning, especially through convolutional neural networks (CNN) such as the U-Net 3D model, has revolutionized fault identification from seismic data, representing a significant leap over traditional methods. Our review traces the evolution of CNN, emphasizing the adaptation and capabilities of the U-Net 3D model in automating seismic fault delineation with unprecedented accuracy. We find: 1) The transition from basic neural networks to sophisticated CNN has enabled remarkable advancements in image recognition, which are directly applicable to analyzing seismic data. The U-Net 3D model, with its innovative architecture, exemplifies this progress by providing a method for detailed and accurate fault detection with reduced manual interpretation bias. 2) The U-Net 3D model has demonstrated its superiority over traditional fault identification methods in several key areas: it has enhanced interpretation accuracy, increased operational efficiency, and reduced the subjectivity of manual methods. 3) Despite these achievements, challenges such as the need for effective data preprocessing, acquisition of high-quality annotated datasets, and achieving model generalization across different geological conditions remain. Future research should therefore focus on developing more complex network architectures and innovative training strategies to refine fault identification performance further. Our findings confirm the transformative potential of deep learning, particularly CNN like the U-Net 3D model, in geosciences, advocating for its broader integration to revolutionize geological exploration and seismic analysis.

APPLICATION OF NEURAL NETWORK FOR THE TRANSFORMER PROTECTIVE RELAYS

A neural network method used to identify the different operating states of transformers has been proposed and established.It is superior to the traditional transformer protective principles and can correctly identify,within half cycle from the fault inception,the internal faults,magnetizing inrush current state,external faults and switching on the internal faults of a no load transformer.In addition,this method has broad availability and high fault tolerant ability.A lot of simulations have demonstrated its superiority.

Parity Relation Based Fault Estimation for Nonlinear Systems: An LMI Approach

This paper proposes a parity relation based fault estimation for a class of nonlinear systems which can be modelled by Takagi-Sugeno （TS） fuzzy models. The design of a parity relation based residual generator is formulated in terms of a family of linear matrix inequalities （LMIs）. A numerical example is provided to illustrate the effectiveness of the proposed design techniques.

傅里叶变换用于地下电缆故障检测(英文)

An analysis of underground power cables is performed using Fourier analysis with the objective of detecting fault and average life of the cables.Three types of cables are used in this experiment:a normal cable,a shorted cable, and a cable with holes.The impedance in each case is computed and Fourier transformation is applied so that the resulting impedance magnitude and impedance phase can be examined in the frequency domain.Various windowing techniques are applied to the experimental data to eliminate any interference.Fourier analysis is then applied to the impedance data calculated from both the sending end voltage and differential voltage.This analysis reveals differences in the frequency response of the three different types of a cable and can eventually be used as a measure for fault detection. Preliminary results reveal the differences in the frequency response.Accordingly,Fourier type methods can be effectively used as low cost and viable solutions to identify and detect faults in underground cables.

Adaptive restarting method for LCC-HVDC based on principle of fault location by current injection

The existing LCC-HVDC transmission project adopts the fixed-time delay restarting method.This method has disadvantages such as non-selectivity,long restart process,and high probability of restart failure.These issues cause a secondary impact on equipment and system power fluctuation.To solve this problem,an adaptive restarting method based on the principle of fault location by current injection is proposed.First,an additional control strategy is proposed to inject a current detection signal.Second,the propagation law of the current signal in the line is analyzed based on the distributed parameter model of transmission line.Finally,a method for identifying fault properties based on the principle of fault location is proposed.The method fully considers the influence of the long-distance transmission line with earth capacitance and overcomes the influence of the increasing effect of the opposite terminal.Simulation results show that the proposed method can accurately identify the fault properties under various complex fault conditions and subsequently realize the adaptive restarting process.