Pattern recognition of partial discharge in XLPE cables using a neural network

Suzuki, H.; Endoh, Takeshi

doi:10.1109/14.142717

Cited by 63 publications

(17 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A 'complete' stochastic characterization of the PD phenomenon requires a determination of a set of many different conditional distributions [15]. In the following discussion, we shall consider as an example the conditional distributions P1(4)iIQ+), i = 1,2,..., where 4>i is the phase-of-occurrence of the i th negative pulse to appear in an arbitrary cycle of the applied alternating voltage, and Q+ is the sum of PD amplitudes that occurred in the corresponding previous positive half cycle, Q+= L qt (1) Again, it is assumed that Q+ is restricted to lie within a specified window. In the case of a point-to-dielectric discharge gap or a cavity in a solid insulator, Q+ is a measure of the net charge deposited on the dielectric surface by the PD events during a given positive half cycle.…”

Section: Quantifying the Effects Of Memory Propagationmentioning

confidence: 99%

“…C ONSIDERABLE interest appears to exist in using information about the statistics of pulsating partialdischarge (PD) phenomena for identification or pattern recognition purposes as evident from numerous recent publications on the subject [1][2][3][4][5]. The types of statistical data considered for pattern recognition are generally the distributions in phase-of-occurrence of PD pulses and the phase-resolved pulse amplitude or average charge distributions [6][7][8][9][10][11].…”

Section: Introductionmentioning

confidence: 99%

“…The purposes of this work are: (1) to show the importance of memory effects using selected results from measurements and computer simulations, (2) to indicate how memory effects are revealed from measurement of conditional distributions,…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Importance of unraveling memory propagation effects in interpreting data on partial discharge statistics

Brunt

Cernyar

Glahn

1993

IEEE Trans. Elect. Insul.

View full text Add to dashboard Cite

The significance of memory propagation in controlling the stochastic behavior of partial-discharge phenomena is demonstrated by determination of various conditional amplitude and phaseof-occurrence distributions for both measured and simulated discharge pulses. A system that can be used to measure directly a set of both conditional and unconditional pulse amplitude and phase distributions needed to reveal memory effects and quantify the phase-resolved stochastic properties of partial-discharge pulses, is briefly described. It is argued that not only is an unraveling of memory effects essential in any attempt to understand the physical basis for the observed stochastic behavior of partial-discharge phenomena, but also that the data on conditional distributions provide additional statistical information that may be needed to optimize the reliability of partial-discharge pattern recognition schemes now being considered for use in insulation testing.

show abstract

Section: Quantifying the Effects Of Memory Propagationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Importance of unraveling memory propagation effects in interpreting data on partial discharge statistics

Brunt

Cernyar

Glahn

1993

IEEE Trans. Elect. Insul.

View full text Add to dashboard Cite

show abstract

“…Using this measurement circuit, the measured data can be the ‘apparent charge’ q , the ‘repetition rate’ n , and the ‘phase amplitude’ φ . These parameters are also used for pattern recognition .…”

Section: Pd Measurement Setupmentioning

confidence: 99%

Data mining on partial discharge signals of power transformer's defect models

Darabad

Vakilian

Blackburn

et al. 2012

Int. Trans. Electr. Energ. Syst.

View full text Add to dashboard Cite

SUMMARY Partial discharge (PD) is a common phenomenon that occurs in the insulation of high‐voltage equipment, such as transformers, and has a damaging effect on the insulation. If data mining techniques are used to extract special specifications and features for different types of PDs in power transformers or other power equipment, one can diagnose the insulation condition of such equipment by its continuous online monitoring. These diagnosis results can be used to develop precise preventive measures and maintenance procedures. This maintenance would cost less and can be carried out in a shorter time. Consequently, the lifetime expectancy of the transformers will be improved. In this article, experimental models are developed to simulate some of the PD types that can occur in a power transformer. Using these models, some of the features that can differentiate those PD types are extracted through texture analysis. Two classification algorithms are used for the separation of PD sources, decision tree and k‐nearest neighbors algorithm. It is anticipated that these models can help us to identify the type of PD fault in a transformer when the required PD measured data are available. Copyright © 2012 John Wiley & Sons, Ltd.

show abstract

“…However, in reality, on site PD measurement suffers from lower detection sensitivity due to the interference of external noises [22]. PD measurement often faces interference caused by radio transmissions, power electronics components, random noise from switching, lightning, arcing, harmonics and interferences from ground connections [23].…”

Section: Introductionmentioning

confidence: 99%

Classification of Partial Discharge Measured under Different Levels of Noise Contamination

2017

View full text Add to dashboard Cite

Cable joint insulation breakdown may cause a huge loss to power companies. Therefore, it is vital to diagnose the insulation quality to detect early signs of insulation failure. It is well known that there is a correlation between Partial discharge (PD) and the insulation quality. Although many works have been done on PD pattern recognition, it is usually performed in a noise free environment. Also, works on PD pattern recognition in actual cable joint are less likely to be found in literature. Therefore, in this work, classifications of actual cable joint defect types from partial discharge data contaminated by noise were performed. Five cross-linked polyethylene (XLPE) cable joints with artificially created defects were prepared based on the defects commonly encountered on site. Three different types of input feature were extracted from the PD pattern under artificially created noisy environment. These include statistical features, fractal features and principal component analysis (PCA) features. These input features were used to train the classifiers to classify each PD defect types. Classifications were performed using three different artificial intelligence classifiers, which include Artificial Neural Networks (ANN), Adaptive Neuro-Fuzzy Inference System (ANFIS) and Support Vector Machine (SVM). It was found that the classification accuracy decreases with higher noise level but PCA features used in SVM and ANN showed the strongest tolerance against noise contamination.

show abstract

Pattern recognition of partial discharge in XLPE cables using a neural network

Cited by 63 publications

References 5 publications

Importance of unraveling memory propagation effects in interpreting data on partial discharge statistics

Importance of unraveling memory propagation effects in interpreting data on partial discharge statistics

Data mining on partial discharge signals of power transformer's defect models

Classification of Partial Discharge Measured under Different Levels of Noise Contamination

Contact Info

Product

Resources

About