• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 1995.04a
• /
• pp.161-164
• /
• 1995

실시간 ANC 이전에, 소음원으로부터 대상계를 거친 신호와 소음원신호가 오차계를 거친 신호를 측정함으로서 이론적인 한계성능을 파악하기 위한 지표를 제시하였다. 이 지표는 오차계를 통과한 신호를 입력으로, 대상계를 통과한 신호를 출력으로 하는 계의 선형화 정도를 나타내는 다중기여도함수와 흡사한 형태로 나타났다. 제한된 상황에서 ANC 성능을 최대화 하고자 할 때 이 지표는 레퍼런스와 부가음원의 위치 및 갯수 결정등 여러가지 파라메터 결정에 매우 유용하게 쓰일 수 있다고 생각된다.

• Computational Structural Engineering
• /
• v.11 no.4
• /
• pp.187-196
• /
• 1998

불연속 갤러킨 정식화에 기초를 둔 시간적분법에 대하여 시간을 변수로 한 유한요소적 접근법을 시도하였다. 단일 형상함수와 두 형상함수 정식화에 대해 각각 선형, 이차 형상함수를 적용하여 모두 네 종류의 시간적분법을 유도하였으며, 각 방법에 대하여 시간시텝의 증가에 따른 변위와 속도의 관계를 나타내는 증폭행렬을 계산하였다. 유도된 방법들의 성능을 평가하기 위하여 부하가 갑자기 변화는 진동 문제를 해석하고 변위의 오차를 비교하였다. 네 가지의 방법에 대하여 국부 오차 추정치를 개발하였으며, 오차 추정치의 정확도를 수치예를 이용하여 평가하였다. 단일 형상함수 정식화에서 이차 형상함수를 이용한 오차 추정치가 실제 국부오차를 잘 나타내었으며 유도된 오차 추정치는 시간간격제어 기법에서 시간간격의 크기를 결정하는 척도로 이용 가능하다.

• Journal of Astronomy and Space Sciences
• /
• v.26 no.1
• /
• pp.111-126
• /
• 2009

We used an optical observation system with a 0.6m wide-field telescope and 5 computers system in KASI (Korean Astronomy and Space Science Institute) for satellite optical observation. Optical data have errors that are caused by targeting, expose start time and end-point determination. Gauss method for initial orbit determination was tested using angle-only data simulated by KODAS. And suitable time span is confirmed for result which has minimum errors. Initial orbit determination results are proved that optical observation system in KASI is possible satellite tracking for a short period. And also through differential correction, initial orbit determination results are improved.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.2
• /
• pp.251-260
• /
• 2017

The option pricing of Black와 Scholes (1973) and Merton (1973) has been widely reported to fail to reflect the time varying volatility of financial time series in many real applications. For example, Duan (1995) proposed GARCH option pricing method through Monte Carlo simulation. However, financial time series is known to follow a fat-tailed and leptokurtic probability distribution, which is not explained by Duan (1995). In this paper, in order to overcome such defects, we proposed the option pricing method based on GARCH models with normal mixture errors. According to the analysis of KOSPI200 option price data, the option pricing based on GARCH models with normal mixture errors outperformed the option pricing based on GARCH models with normal errors in the unstable period with high volatility.

• The Journal of the Korea Contents Association
• /
• v.10 no.5
• /
• pp.99-106
• /
• 2010

Multilayer perceptrons(MLPs) or feed-forward neural networks are widely applied to many areas based on their function approximation capabilities. When implementing MLPs for application problems, we should determine various parameters and training methods. In this paper, we discuss the design of MLPs especially for pattern classification problems. This discussion includes how to decide the number of nodes in each layer, how to initialize the weights of MLPs, how to train MLPs among various error functions, the imbalanced data problems, and deep architecture.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2010.05a
• /
• pp.625-629
• /
• 2010

암거 수리설계는 상류부의 수위를 과다하게 상승시키지 않는 상태에서 안전하게 계획홍수량을 하류로 소통시킬 수 있도록 최적의 단면을 결정하는 것이다. 현재 국내 암거 설계는 한국도로공사(1991)에서 발표한 "도로배수계획"을 이용하고 있는데, 이는 주로 미국도로성(FHWA)에서 발표한 설계기법을 따른 것이며, 이 기법은 설계항목과 관련되는 설계유량과 지배단면을 결정하여 시행오차적인 방법으로 계획 상류부 수심을 산정하는 순서로 진행된다. 상류부 수심을 결정하기 위해서 미국도로성의 기법은 계산도표를 이용하여 손실 수두와 한계수심을 산정한다. 그러나 계산도표를 이용하여 설계할 경우 주관적인 오차가 생길 수 있으며 전산해석에 있어서도 큰 장애요인이 될 수 있다. 본 연구에서는 기존 연구에서 기술된 암거 설계법을 전반적으로 재검토하여 일부 수식 전개 과정에서 발생한 오류를 수정하고, 암거의 규격 산정 시 암거의 형상을 미리 결정하여 암거 프로그램 개발 시 효율적일 것으로 판단되는 새로운 방법을 개발하였다.

• Journal of Korea Society of Industrial Information Systems
• /
• v.18 no.1
• /
• pp.81-89
• /
• 2013

The purpose of this study is to analyze the information effect about the pricing using the ETF price, the benchmark index, and the total tracking error between the ETF price and the benchmark index on the index ETF market and sector ETF markets. Furthermore, the total tracking error is distinguished between the market tracking error and the NAV tracking error. Summary of this study are as follows: First, While KODEX200 don't have impact factors on the price, the most sectors of ETF have the factors affecting the pricing decision. They are the day before the total tracking error or market tracking error. Second, for the ETF price of the most industry, we find that the day before the market tracking error have the price discovery function because it is a negative(-) coefficients. But NAV tracking error could not find such a feature. Finally, the sector ETF price of energy chemical, construction, IT, and semiconductor industries affected of the day before positive(+) impact by the benchmark index price.

• Journal of Advanced Navigation Technology
• /
• v.24 no.2
• /
• pp.92-100
• /
• 2020

Satellite-based argumentation systems (SBAS) is a system that enhances the accuracy, integrity, availability and continuity of GNSS navigation users by using geostationary orbit (GEO) satellites to send correction information and the failures of global navigation satellite system (GNSS) satellites in the form of messages. The correction information provided by SBAS is pseudorange error, satellite orbit error, clock error, and ionospheric delay error at 250 bps. Therefore, A lot of message processing are required for the SBAS navigation. There is a need to reduce SBAS time to first fix (TTFF) for using SBAS navigation in systems with short operating time. In this paper, A-SGNSS operation method was proposed for reducing SBAS TTFF. Also, A-SGNSS TTFF and availability were analyzed.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.11 no.4
• /
• pp.325-330
• /
• 2018

This paper, following the shape parameters of the minimax distribution, describes the special form of the beta distribution, the Minimax distribution, as a function of the shape parameters for the software reliability model based on the non-homogeneous Poisson process. Characteristics and usefulness were discussed. As a result, the case of the shape parameter 1 of Minimax distribution than less than and greate in mean squared error is the smallest, in determination coefficient, appears to be high, the shape parameter 1 of Minimax distribution regard as an efficient model. The estimated determination coefficient of the proposed model is estimated to be more than 95%, which is a useful model in the field of software reliability. Through this study, software design and users can identify the software failure characteristics using mean square error, decision coefficient, and confidence interval can be used as a basic guideline.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.16 no.3
• /
• pp.89-95
• /
• 2016

In this paper, we propose a dual-mode blind equalization algorithm that two of the blind equalization algorithm using the size of the decision-directed error signal is automatically switched. The proposed algorithm has a faster convergence speed due to operation of the MSAGF-SMMA with large fixed step-size mainly in the initial equalization. After the equalization has been made to some extent, the proposed algorithm has a smaller residual error in the steady- state by operation of the MSAGF-SMMA with a variable step-size mainly. The variable step-size is determined by multiplying the size of the decision-directed error signal of a fixed step-size. In this paper, we analyze the performance of the proposed algorithm. The computer simulation results demonstrate that the proposed algorithm has a significantly improved performance in terms of a residual inter-symbol interference and residual error in the steady-state compared with the MMA, SMMA, and MSAGF-SMMA.