• 한국태양에너지학회:학술대회논문집
• /
• 2008.11a
• /
• pp.81-84
• /
• 2008

This paper discussed the validity of wind force power generation in consideration of the topographical characteristics of Korea. In order to estimate the exact generation of wind power plants, we analyzed and compared wind resources in mountain areas and plain areas by introducing not only wind velocity, the most important variable, but also wind distribution and wind standard deviation that can reflect the influence of landform sufficiently. According to the results of this study, generation was higher at wind power plants installed in southwestern coastal areas where wind velocity was low than at those installed in mountain areas in Gangwondo where wind velocity was high. This suggests that the shape parameter of wind distribution is low due to the characteristics of mountain areas. and the standard deviation of wind velocity is large due to the effect of mountain winds, and therefore, actual generation is low in mountain areas although wind velocity is high.

• The Journal of the Acoustical Society of Korea
• /
• v.23 no.8
• /
• pp.590-596
• /
• 2004

This paper Presents a new self-adaptation algorithm based on maximum a posteriori (MAP) eigenvoice for Korean connected digit recognition. The proposed MAP eigenvoice is developed by introducing a probability density model for the eigenvoice coefficients. The Proposed approach provides a unified framework that incorporates the Prior model into the conventional eigenvoice estimation. In self-adaptation system we use only one adaptation utterance that will be recognized, we use MAP eigenvoice that is most robust adaptation. In series of self-adaptation experiments on the Korean connected digit recognition task. we demonstrate that the performance of the proposed approach is better than that of the conventional eigenvoice algorithm for a small amount of adaptation data.

• Journal of the Korean Solar Energy Society
• /
• v.28 no.6
• /
• pp.24-32
• /
• 2008

This paper discussed the Feasibility study of wind power generation considering the topographical characteristics of Korea. In order to estimate the exact generation of wind power plants, we analyzed and compared wind resources in mountain areas and plain areas by introducing not only wind speed, the most important variable, but also wind distribution and wind standard deviation that can reflect the influence of landform sufficiently. According to the results of this study, generation was almost the same at wind power plants installed in southwestern coastal areas where wind speed was low as at those installed in mountain areas in Gangwondo where wind speed was high. This demonstrates that the shape parameter of wind distribution is low due to the characteristics of mountain areas, and the standard deviation of wind speed is large due to the effect of mountain winds, therefore, actual generation compared to southwestern coastal areas is almost similar in mountain areas even though wind speed is high.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2012.01a
• /
• pp.41-43
• /
• 2012

Today many people are greatly interested in the environment. Especially increasing affluent of food causes a great amount of food waste. To handle this effectively, we now have a lot of problems of disposing garbage all over the world. In Korea, in order to reduce this garbage, we should use a standard plastic garbage bag in which we have to throw away our garbage. So it has an effect on significantly reducing the waste amount every year. Now, there are a lot of cases that residents again use a standard plastic garbage bag. The purpose of this study is to propose algorithm preventing the re-cycling of a standard plastic garbage bag.

• Food Science and Preservation
• /
• v.7 no.4
• /
• pp.357-361
• /
• 2000

Watercore influences storage and distribution industry of fruit. Therefore, the technique for nondestructive discrimination of watercore fruit is needed. This work reports about the possibility of watercore discrimination of ante using transmitted light, and the effects of various factors. CCD camera was used to capone images of each apple fruit. An excess watercore apples were higher light transmission score than little watercore apples. The accuracy fur discrimination of watercore apple was about 70% using transmitted light. Peel thickness, anthocyanin layer thickness and density of apple affected the light transmission. Apples having thin peel, thin pigment layer and low density tended to high transmitted light score. Apples of good color degree were more probability of existence watercore than ones of bad color degree. But color distribution of apple peel was not correlated with watercore.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.39 no.2
• /
• pp.147-154
• /
• 2002

In this paper, we propose an error estimation method based on the Bhattacharyya distance for multimodal data. First, we try to find the empirical relationship between the classification error and the Bhattacharyya distance. Then, we investigate the possibility to derive the error estimation equation based on the Bhattacharyya distance for multimodal data. We assume that the distribution of multimodal data can be approximated as a mixture of several Gaussian distributions. Experimental results with remotely sensed data showed that there exist strong relationships between the Bhattacharyya distance and the classification error and that it is possible to predict the classification error using the Bhattacharyya distance for multimodal data.

• Annual Conference on Human and Language Technology
• /
• 2021.10a
• /
• pp.340-345
• /
• 2021

최근 많은 연구들이 BERT를 활용하여, 주어진 문맥에서 언어학/문법적으로 적절하지 않은 단어를 인지하고 찾아내는 성과를 보고하였다. 하지만 일반적으로 딥러닝 관점에서 NLL기법(Negative log-likelihood)은 주어진 문맥에서 언어 변칙에 대한 정확한 성격을 규명하기에는 어려움이 있다고 지적되고 있다. 이러한 한계를 해결하기 위하여, Li et al.(2021)은 트랜스포머 언어모델의 은닉층별 밀도 추정(density estimation)을 통한 가우시안 확률 분포를 활용하는 가우시안 혼합 모델(Gaussian Mixture Model)을 적용하였다. 그들은 트랜스포머 언어모델이 언어 변칙 예문들의 종류에 따라 상이한 메커니즘을 사용하여 처리한다는 점을 보고하였다. 이 선행 연구를 받아들여 본 연구에서는 한국어 기반 언어모델인 KoBERT나 KR-BERT도 과연 한국어의 상이한 유형의 언어 변칙 예문들을 다른 방식으로 처리할 수 있는지를 규명하고자 한다. 이를 위해, 본 연구에서는 한국어 형태통사적 그리고 의미적 변칙 예문들을 구성하였고, 이 예문들을 바탕으로 한국어 기반 모델들의 성능을 놀라움-갭(surprisal gap) 점수를 계산하여 평가하였다. 본 논문에서는 한국어 기반 모델들도 의미적 변칙 예문을 처리할 때보다 형태통사적 변칙 예문을 처리할 때 상대적으로 보다 더 높은 놀라움-갭 점수를 보여주고 있음을 발견하였다. 즉, 상이한 종류의 언어 변칙 예문들을 처리하기 위하여 다른 메커니즘을 활용하고 있음을 보였다.

• Journal of the Korean Vacuum Society
• /
• v.16 no.6
• /
• pp.489-495
• /
• 2007

The assembly of the main system of the KSTAR tokamak has been recently completed, and the preparation for the 1st plasma and test operations is progressed. The fueling system established for these purposes uses only one port placed at the opposite side of the pumping duct, and has a difficulty of attaining a uniform and fast supply of fuel particles to the plasma. At the base operation stage after finishing the test operation, the fueling system must be improved to provide a uniform fueling and a feed-back control in accordance with a high-density tokamak plasma maintained for a long period. As a part for understanding the points to be improved in the fueling system, a Monte Carlo simulation on the gas fueling into the tokamak plasma has been executed. After modeling the vacuum vessel and the plasma of quasi-D shapes as tori of rectangular cross-sections, the influences of the position and the number of the fueling inputs on the particle density distribution for a given pumping probability and mean free path were investigated.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.53 no.10
• /
• pp.3-9
• /
• 2016

Current mobile communication systems utilize the multiple-input multiple-output (MIMO) transmission technique as an important means to enhance the bandwidth efficiency. Accurate beamforming via channel estimation contributes to the signal-to-interference-plus-noise ratio (SINR) increase and the system performance improvement when MIMO transmission techniques are employed. Therefore, determination of beamforming vectors as well as the design of appropriate codebooks defining these codevectors play an important role in system operation. In this paper, we statistically analyze the phase difference between the channels corresponding to adjacent antenna elements in order to design an efficient codebook for uniform circular arrays (UCAs). We introduce new parameters which compensate for the additional phase difference observed in its probability density functions (PDFs). The performance of the proposed codebook is tested using the spatial channel model (SCM) to demonstrate its gain over the standard codebooks adopted in the long term evolution (LTE) Releases 8 and 10.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.29 no.3
• /
• pp.233-240
• /
• 2018

Radar cross section (RCS) analysis of chaff clouds is essential for the accurate detection and tracking of missile targets using radar. For this purpose, we compare the performance of two existing methods of predicting RCS of chaff clouds. One method involves summing up the RCS values of individual chaffs in a cloud, while the other method predicts the RCS values using aerodynamic models based on the probability density function. In order to compare and analyze the two techniques more precisely, the RCS of a single chaff computer-aided design model consisting of a half wavelength dipole was calculated using the commercial electromagnetic numerical analysis software, FEKO 7.0, to estimate the RCS values of chaff clouds via simulation. Thus, we verified that our method using the probability density distribution model is capable of analyzing the RCS of chaff clouds more efficiently.