• 대한전기학회논문지
• /
• 제43권4호
• /
• pp.594-600
• /
• 1994

The maximum charge accumulation in transformers caused by streaming electrification is commonly observed on the bottom area of transformer winding(i.e.the oil inlet of transformer winding). It is because turbulent flows occur in this portion of transformers. Consequently, if we are to reduce the maximum charge accumulation of transformers, it is important that we should have a good understanding of the conditions of the occurence of partial turbulent flows in the oil inlet of transformer winding. With this point in mind, a simple flow model was designed in this study in order to simulate oil flow in the insulation ducts of power transformers. As a method for mitigating charge accummulation, attention was given to (a) a flow path design for the improvement of flow conditions, and (b) the charge tendency of the material used this design. The results of this study show that the above method is useful for mitigating charge accumulation.

• 대한전기학회논문지:전력기술부문A
• /
• 제52권8호
• /
• pp.463-470
• /
• 2003

This paper proposes an expert system, which is able to determine the problem-solving strategy for diversity events occurred on the distribution system. The events include events related to the fault, scheduled outage and optimal operating task such as feeder line fault, line scheduled outage, line overload, system load balancing, system loss minimization, main transformer fault, main transformer scheduled outage, main transformer overload, main transformer protection control. The expert system enhances the reliability of software designed by the integrated concept for the diversity events. The expert system is implemented in C language. And the effectiveness and accuracy for the expert system is verified by simulating the event cases for typical distribution model.

• 대한전기학회논문지:전기기기및에너지변환시스템부문B
• /
• 제48권10호
• /
• pp.523-529
• /
• 1999

In the case of the large power transformer, the grain-oriented material is usually used. So, to obtain more accurate results, anisotropy and non-linearity of the material must be considered. The Newton-Raphson(NR) method is generally used for analyzing these non-linear properties, but it consumes so much time, especially when the number of nodes is large or the shape of the model is complex. The transmission line modeling (TLM) method is successfully adopted to the analysis of non-linear properties with FEM, but it has not been adopted to the analysis of the anisotropic material. In this paper, the formulation of the TLM method considering anisotropy is developed and the adoption to the 3-phase transformer is presented.

• 전기학회논문지
• /
• 제60권2호
• /
• pp.301-308
• /
• 2011

This paper deals with the coupled electromagnetic-mechanical field analysis for short-circuit electromagnetic force of the dry-type transformer. The short-circuit currents are calculated using external circuit in accordance with short-circuit test equipment. According to short-circuit current, the generated magnetic leakage flux density in dry-type transformer model is calculated by finite element method. The radially-directed electromagnetic forces in windings are calculated using electromagnetic field analysis and then axially-directed electromagnetic forces in windings are calculated using electromagnetic-mechanical field analysis. The calculated axially-directed electromagnetic forces in high voltage winding are compared to those of measured ones and showed good agreement with experimental results.

• 대한전기학회논문지:전력기술부문A
• /
• 제50권4호
• /
• pp.157-165
• /
• 2001

The two criteria to identify the disturbances of the power transformer has been reported in this paper. They have been derived through EMTP simulations of internal faults, inrush and overexcitation for the model of 154/22.9[kV], 40[MVA], Y-Y three-phase power transformer. We propose the crisp algorithm which uses two criteria. A series of test results clearly indicate that the method can identify not only an internal fault but also the other transients. The average of relay operation times is about 7.2[ms]. The proposed algorithm immunes to the transient state.

• 전기학회논문지P
• /
• 제55권2호
• /
• pp.110-114
• /
• 2006

This paper describes the accident analysis by modelling the current transformer mounting in a metal-clad switchgear at 22.9kV. In analyzing the accident, the reconstruction at the current transformer mounting(VCB connecting guide) has to be taken into account. The accident was modelled as a 3-phase ground fault occurring between the end plate of 22.9kV lines and the safety shutter at the current transformer mounting of the VCB inside the metal clad switchgear. Since the outside maintenance of the metal clad switchgear is restricted by the enclosed compartments, Its circumference has to be kept clean. Through the reconstruction results, it was confirmed that the fault of the enclosed switchboard could be reduced when the shutter made of Fe material was chanted into an insulation.

• 대한임베디드공학회논문지
• /
• 제17권6호
• /
• pp.319-327
• /
• 2022

Transformers are the most famous deep learning models that has achieved great success in natural language processing and also showed good performance on computer vision. In this survey, we categorized transformer-based models for computer vision, particularly object detection tasks and perform comprehensive comparative experiments to understand the characteristics of each model. Next, we evaluated the models subdivided into standard transformer, with key point attention, and adding attention with coordinates by performance comparison in terms of object detection accuracy and real-time performance. For performance comparison, we used two metrics: frame per second (FPS) and mean average precision (mAP). Finally, we confirmed the trends and relationships related to the detection and real-time performance of objects in several transformer models using various experiments.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2023년도 추계학술발표대회
• /
• pp.557-560
• /
• 2023

In this paper, we introduce a pre-training method leveraging the capabilities of the Vision Transformer (ViT) for disease diagnosis in conventional Fundus images. Recognizing the need for effective representation learning in medical images, our method combines the Vision Transformer with a Masked Autoencoder to generate meaningful and pertinent image augmentations. During pre-training, the Masked Autoencoder produces an altered version of the original image, which serves as a positive pair. The Vision Transformer then employs contrastive learning techniques with this image pair to refine its weight parameters. Our experiments demonstrate that this dual-model approach harnesses the strengths of both the ViT and the Masked Autoencoder, resulting in robust and clinically relevant feature embeddings. Preliminary results suggest significant improvements in diagnostic accuracy, underscoring the potential of our methodology in enhancing automated disease diagnosis in fundus imaging.

• 한국컴퓨터정보학회:학술대회논문집
• /
• 한국컴퓨터정보학회 2024년도 제69차 동계학술대회논문집 32권1호
• /
• pp.35-36
• /
• 2024

전신 마취 수술 중 저혈압의 발생은 다양한 합병증을 유발하며 이를 사전에 예측하여 대응하는 것은 매우 중요한 일이다. 따라서 본 연구에서는 SHAP 모델을 통해 변수 선택을 진행하고, Transformer 모델을 이용해 저혈압 발생 여부를 예측함으로써 임상적 의사결정을 지원한다. 또한 기존 연구들과는 달리, 수술실에서 수집되는 데이터를 기반으로 하여 높은 범용성을 가진다. 비침습적 혈압 예측에서 RMSE 9.46, MAPE 4.4%를 달성하였고, 저혈압 여부를 예측에서는 저혈압 기준 F1-Score 0.75로 우수한 결과를 얻었다.

• 지식경영연구
• /
• 제22권4호
• /
• pp.27-40
• /
• 2021

Click-Through Rate(CTR) 예측은 추천시스템에서 후보 항목의 순위를 결정하고 높은 순위의 항목들을 추천하여 고객의 정보 과부하를 줄임과 동시에 판매 촉진을 통한 수익 극대화를 달성할 수 있는 핵심 기능이다. 자연어 처리와 이미지 분류 분야는 심층신경망(deep neural network)의 활용을 통한 괄목한 성장을 하고 있다. 최근 이 분야의 주류를 이루던 모델과 차별화된 어텐션(attention) 메커니즘 기반의 트랜스포머(transformer) 모델이 제안되어 state-of-the-art를 달성하였다. 본 연구에서는 CTR 예측을 위한 트랜스포머 기반 모델의 성능 향상 방안을 제시한다. 자연어와 이미지 데이터와는 다른 이산적(discrete)이며 범주적(categorical)인 CTR 데이터 특성이 모델 성능에 미치는 영향력을 분석하기 위해 임베딩의 일반화(regularization)와 트랜스포머의 정규화(normalization)에 관한 실험을 수행한다. 실험 결과에 따르면, CTR 데이터 입력 처리를 위한 임베딩 과정에서 L2 일반화의 적용과 트랜스포머 모델의 기본 정규화 방법인 레이어 정규화 대신 배치 정규화를 적용할 때 예측 성능이 크게 향상됨을 확인하였다.