High impedance fault detection using combination of multi-layer perceptron neural networks based on multi-resolution morphological gradient features of current waveform

Sarlak, M.; Shahrtash, S. Mohammad

doi:10.1049/iet-gtd.2010.0702

Cited by 93 publications

(39 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Table 9 demonstrates the comparison between the proposed stransform based technique, and existing/proposed HIF detection methods. The information related to SNR 30 dB case are not available in the reference paper [5,12,29,30] and is shown as NA (not available) in Table 9. The table clearly reveals an inherent performance of the technique to detect high impedance fault under noisy condition…”

Section: B Results Svm Classifier Using Input As St Featurementioning

confidence: 99%

A Universal High Impedance Fault Detection Technique for Distribution System Using S-Transform and Pattern Recognition

Mishra

Routray

Rout

2016

Technol Econ Smart Grids Sustain Energy

View full text Add to dashboard Cite

This paper describes a new method for high impedance fault (HIF) detection based on s-transform (ST) and pattern recognition technique. Conventional distance, over current and ground fault relays are difficult to apply for High Impedance Fault (HIF) detection in distribution line because of sensitivity, diversity, selectivity issues in case of low value of fault current. Recently, s-transform has been successfully applied for different power system protection problem. It is a very useful tool to analyze transient fault signal to provide both time and frequency information unlike Fourier transform and the same has been considered for high impedance fault detection in this work. The features extracted using s-transform to train and test the two different intelligent classifier like artificial neural network (ANN) and support vector machine (SVM) separately, to discriminate the HIF with other transient phenomenon (Load switching, capacitor Switching) and also normal fault condition. A comparative study of these two classifiers has been reported based on their detection accuracy. It has been found that the proposed techniques are highly effective for high impedance fault detection under a wide range of operating conditions and noisy environment in a high voltage distribution

show abstract

Section: B Results Svm Classifier Using Input As St Featurementioning

confidence: 99%

A Universal High Impedance Fault Detection Technique for Distribution System Using S-Transform and Pattern Recognition

Mishra

Routray

Rout

2016

Technol Econ Smart Grids Sustain Energy

View full text Add to dashboard Cite

show abstract

“…Note: In Step 1, the cut-off frequency of 500 Hz is chosen owing to the fact that low order harmonics, such as 2 nd harmonic and 3 rd harmonic are most evident in fault current [5][6][7]. A filter with such a cut-off frequency can eliminate noises and at the same time reserve the useful low order harmonics.…”

Section: Detection Algorithm Based On CCC Of Zerosequence Currentmentioning

confidence: 99%

“…Many fault features have been extracted and the highly recognized ones are: radiation behavior [1][2][3] and harmonic distortions in voltage and current [4][5][6][7][8][9]. Plenty of detection algorithms have been proposed and analyzed, including electromagnetic radiation based algorithms [3], harmonic based algorithms [4][5][6][7][8][9][10][11][12][13][14], wavelet based algorithms [15][16][17], instantaneous power based algorithms [18], and some intelligent detection algorithms [19,20]. Due to the obvious 3 rd harmonic characteristic of HIFs, the harmonic based algorithms are the most commonly adopted in industrial application.…”

Section: Introductionmentioning

confidence: 99%

Arc flash fault detection in wind farm collection feeders based on current waveform analysis

Wang

Geng

et al. 2016

J. Mod. Power Syst. Clean Energy

View full text Add to dashboard Cite

High impedance faults (HIFs) are easy to occur in collective feeders in wind farms and may cause the cascading of wind generators tripping. This kind of faults is difficult to be detected by traditional relay or fuse due to the limited fault current values and the situation is worse in wind farms. The mostly adopted HIF detection algorithms are based on the 3 rd harmonic characteristic of the fault zero-sequence currents, whereas these 3 rd harmonics are very easy to be polluted by wind power back-to-back converters. In response to this problem, the typical harmonic characteristic of HIF arc flash based on Mayr's arc model is first analyzed, and the typical fault waveforms of HIF in wind farm are presented. Then the performance of the harmonic based HIF detection algorithm is discussed, and a novel detection algorithm is proposed from the viewpoint of time domain, focusing on the convex and concave characteristic of zero-sequence current at zerocrossing points. A HIFs detection (HIFD) prototype implementing the proposed algorithm has been developed. The sensitivity and security of the algorithm are proved by field data and RTDS experiments.

show abstract

“…Sarlak and Shahrtash [6] developed a pattern recognition-based algorithm for detecting HIFs with broken and unbroken conductors and distinguishing them from other similar phenomena. This method employed multi-resolution morphological gradient for extraction of the time based features, and according to these features, three multilayer perceptron neural networks are trained and the outputs are combined.…”

Section: Introductionmentioning

confidence: 99%

Voltage Based Detection Method for High Impedance Fault in a Distribution System

Thomas

Bhaskar

Prakash³

2015

J. Inst. Eng. India Ser. B

View full text Add to dashboard Cite

High-impedance faults (HIFs) on distribution feeders cannot be detected by conventional protection schemes, as HIFs are characterized by their low fault current level and waveform distortion due to the nonlinearity of the ground return path. This paper proposes a method to identify the HIFs in distribution system and isolate the faulty section, to reduce downtime. This method is based on voltage measurements along the distribution feeder and utilizes the sequence components of the voltages. Three models of high impedance faults have been considered and source side and load side breaking of the conductor have been studied in this work to capture a wide range of scenarios. The effect of neutral grounding of the source side transformer is also accounted in this study. The results show that the algorithm detects the HIFs accurately and rapidly. Thus, the faulty section can be isolated and service can be restored to the rest of the consumers.

show abstract

High impedance fault detection using combination of multi-layer perceptron neural networks based on multi-resolution morphological gradient features of current waveform

Cited by 93 publications

References 20 publications

A Universal High Impedance Fault Detection Technique for Distribution System Using S-Transform and Pattern Recognition

A Universal High Impedance Fault Detection Technique for Distribution System Using S-Transform and Pattern Recognition

Arc flash fault detection in wind farm collection feeders based on current waveform analysis

Voltage Based Detection Method for High Impedance Fault in a Distribution System

Contact Info

Product

Resources

About