• 전기학회논문지
• /
• 제67권1호
• /
• pp.114-123
• /
• 2018

This paper concerns a design and learning method of softmax function neural networks based on K-means clustering. The partial discharge data Information is preliminarily processed through simulation using an Epoxy Mica Coupling sensor and an internal Phase Resolved Partial Discharge Analysis algorithm. The obtained information is processed according to the characteristics of the pattern using a Motor Insulation Monitoring System program. At this time, the processed data are total 4 types that void discharge, corona discharge, surface discharge and slot discharge. The partial discharge data with high dimensional input variables are secondarily processed by principal component analysis method and reduced with keeping the characteristics of pattern as low dimensional input variables. And therefore, the pattern classifier processing speed exhibits improved effects. In addition, in the process of extracting the partial discharge data through the MIMS program, the magnitude of amplitude is divided into the maximum value and the average value, and two pattern characteristics are set and compared and analyzed. In the first half of the proposed partial discharge pattern classifier, the input and hidden layers are classified by using the K-means clustering method and the output of the hidden layer is obtained. In the latter part, the cross entropy error function is used for parameter learning between the hidden layer and the output layer. The final output layer is output as a normalized probability value between 0 and 1 using the softmax function. The advantage of using the softmax function is that it allows access and application of multiple class problems and stochastic interpretation. First of all, there is an advantage that one output value affects the remaining output value and its accompanying learning is accelerated. Also, to solve the overfitting problem, L2-normalization is applied. To prove the superiority of the proposed pattern classifier, we compare and analyze the classification rate with conventional radial basis function neural networks.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2004년도 하계학술대회 논문집 C
• /
• pp.1982-1984
• /
• 2004

Stator coil of rotating machinery has shown different characteristics according to impregnated with coil or not. And this is major determinant of equipment's life. In this paper, PD characteristics is studied as a classification scheme between two specimens. Processing of the coil impregnation is very important thing because that influences on thermal and electrical characteristics of the coil. And then PD is occurring at the coil and causing insulation degradation. For processing statistical processing, PD data acquired from PD detector using PDASDA(partial discharge acquisition, storage and display system). And also these statistical distribution and parameter are applied to classify PD sources by neural networks. As a result of, Neural Networks have a good discrimination rate for classification PD sources.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1995년도 하계학술대회 논문집 C
• /
• pp.1293-1295
• /
• 1995

This paper describes the detection of the ultrasonic signals reduced by barrier in test transformer. The ultrasonic signals are generated by partial discharges which cause the insulation failure of transformer. The ultrasonic signals are reduced by barrier. But it was shown that the reduced ultrasonic signals can be measured as the location of the ultrasonic signal detectors is selected properly.

• International Journal of Advanced Culture Technology
• /
• 제4권1호
• /
• pp.31-36
• /
• 2016

Morphology evolution of the active layer in bulk hetero-junction organic photovoltaic is modeled and visualized. The width of the phase domain can be predicted using the relationship of characteristics length and evolution time of the process. The 3D numerical simulation of the PCBM/P3HT blend morphology evolution with respect to time is presented. It is observed that the domain width of composition phase can be predicted by using the relationship between value of characteristic length R(t) and evolution time t.

• 한국전기전자재료학회:학술대회논문집
• /
• 한국전기전자재료학회 2009년도 추계학술대회 논문집
• /
• pp.228-228
• /
• 2009

In stator windings of a 6.6kV rotating machine, internal discharge, corona discharge, and surface discharge are mainly caused by internal voids and insulation degradation. In the case of the discharge spark and electromagnetic pulse generated from a discharge source, we can detect it using various RF resonators like an EM sensor. In order to detect these types of electromagnetic sources, we designed and fabricated a planar patch sensor using a CST-MWS simulation, and PD signals from artificial defected stator windings were also measured by the sensor proposed in this study. Furthermore, In the results of the experiment, it showed similar performance to the HFCT sensor.

• 한국초전도저온공학회:학술대회논문집
• /
• 한국초전도저온공학회 2002년도 학술대회 논문집
• /
• pp.5-9
• /
• 2002

In this paper, a relationship between AC discharge progress and the radiated electromagnetic waves was investigated by measuring electromagnetic waves using a biconical antenna and a spectrum analyzer at the atmospheric pressure in L$N_{2}$ . From the results of this study, according to the consideration of the mutual relation between frequency spectrum of the radiated electromagnetic waves and discharge progress, it was confirmed that detecting partial discharge and estimating discharge progress can be possible. It Is considered that these results obtained from this investigation may be used as the fundamental data for diagnosis and prediction of insulation on superconducting and cryogenic power equipments.

• 한국전기전자재료학회논문지
• /
• 제19권10호
• /
• pp.976-983
• /
• 2006

Multi-resolution Signal Decomposition (MSD) Technique of Wavelet Transform has interesting properties of capturing the embedded horizontal, vertical and diagonal variations within an image in a separable form. This feature was exploited to identify individual partial discharge sources present in multi-source PD pattern, usually encountered during practical PD measurement. Employing the Daubechies wavelet, feature were extracted from the third level decomposed and reconstructed horizontal and vertical component images. These features were found to contain the necessary discriminating information corresponding to the individual PD sources and multi-PD soruces.

• 한국전기전자재료학회논문지
• /
• 제20권1호
• /
• pp.65-73
• /
• 2007

Pattern classification of single and multiple discharge sources was applied using a wavelet image signal method in which a feature extraction was applied using a hidden sub-image. A feature extracting method that used vertical and horizontal images using an MSD method was applied to an averaging process for the scale of pulses for the phase. A feature extracting process for the preprocessing of the input of a neural network was performed using an inverse transformation of the horizontal, vertical, and diagonal sub-images. A back propagation algorithm in a neural network was used to classify defective signals. An algorithm for wavelet image processing was developed. In addition, the defective signal was classified using the extracted value that was quantified for the input of a neural network.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2011년도 제42회 하계학술대회
• /
• pp.2130-2131
• /
• 2011

Electric fire breaks out by various causes such as electric short, overload(over current), leakage current, poor contact, static electricity, insulation deterioration, partial disconnection and etc. In this paper, we studied analysis technique for scientific and systematic judgment for distinction of electricity fire existence and nonexistence, and described about electricity fire cause analysis process for collecting physical evidences. We handled external form analysis, Substance extension analysis, X-ray transmission analysis, cross-section analysis, thermal analysis, chemical structure analysis, surface structure analysis, component analysis, electrical properties analysis, and circuit analysis by analysis technique for electricity fire cause judgment. We expect that analysis technic elevation and correct cause of fire analysis of scene of a fire investigators and analysts through this paper which see help.

• 한국분말재료학회지
• /
• 제21권5호
• /
• pp.389-398
• /
• 2014

Macroporous ceramics with tailored pore size and shape could be used for well-established and emerging applications, such as molten metal filtration, biomaterial, catalysis, thermal insulation, hot gas filtration and diesel particulate filters. In these applications, unique properties of porous materials were required which could be achieved through the incorporation of macro-pores into ceramics. In this article, we reviewed the main processing techniques which can be used for the fabrication of macroporous ceramics with tailored microstructure. Partial sintering, replica templates, sacrificial fugutives, and direct foaming techniques was described here and compared in terms of microstructures and mechanical properties that could be achieved. The main focus was given to the direct foaming technique which was simple and versatile approach that allowed the fabrication of macro-porous ceramics with tailored features and properties.