• 지구물리와물리탐사
• /
• 제25권4호
• /
• pp.242-251
• /
• 2022

자동 미분은 컴퓨터를 이용한 미분 계산시 함수의 전방향 연산만 정의하면 연쇄법칙을 이용해 함수의 미분을 자동으로 계산해주는 기능이다. 최근 자동 미분을 지원하는 계산 라이브러리들의 발달에 따라 지구 물리 역산 문제에 자동 미분 기술을 도입하는 사례가 증가하고 있다. 본 연구에서는 탄성파 탐사 전파형 역산 목적함수의 그래디언트 계산시 자동 미분을 도입하였을 때의 장단점과 성능을 분석하였다. 목적함수의 그래디언트 계산은 연쇄법칙을 이용해 매개변수 미분, 파동장 미분과 목적함수 미분의 곱으로 나타낼 수 있다. 수치 예제를 통해 매개변수 미분과 목적함수 미분 계산은 자동 미분으로 수행하고 파동장 미분은 해석적 방법으로 수행함으로써 해석적 미분의 장점인 빠른 성능과 자동 미분의 장점인 복잡한 그래디언트 계산의 편리성을 함께 얻을 수 있음을 보였다.

• 한국지구물리탐사학회:학술대회논문집
• /
• 한국지구물리탐사학회 2003년도 Proceedings of the international symposium on the fusion technology
• /
• pp.459-463
• /
• 2003

Waveform inversion requires extracting a reliable low frequency content of seismic data for estimating of the low wave number velocity model. The low frequency content of the seismic data is usually discarded or neglected because of the band-limited response of the source and the receivers. In this study, however small the spectral of the low frequency seismic data is, we assume that it is possible to extract a reliable phase information of the low frequency from the seismic data and use it in waveform inversion. To this end, we exploit the frequency domain finite element modeling and source-receiver reciprocity to calculate the $Frech\`{e}t$ derivative of the phase of the seismic data with respect to the earth model parameter such as velocity, and then apply a damped least squares method to invert the phase of the seismic data. Through numerical example, we will attempt to demonstrate the feasibility of our method in estimating the correct velocity model for prestack depth migration.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2015년도 제46회 하계학술대회
• /
• pp.772-773
• /
• 2015

This paper presents a dynamic model of 2-D magnetostriction in electrical steel sheet (ESS) under rotating flux magnetization conditions and its implementation in finite element method (FEM). For an arbitrary waveform of magnetic flux density (B), the corresponding magnetostriction waveform can be predicted by the model. In order to apply the model to FEM easily, the model is based on trilinear interpolation method. As an example, the model is applied to a three-phase transformer constructed by highly grain-oriented electrical steel sheets and the numerical results by the magnetostriction model are discussed.

• 한국정보디스플레이학회:학술대회논문집
• /
• 한국정보디스플레이학회 2002년도 International Meeting on Information Display
• /
• pp.423-426
• /
• 2002

New driving waveform which can be inferred by analyzing fundamental LC mechanism is suggested. The main idea of the new waveform is the stabilizing LC layer fast in the cell by controlling the middle director of LC layer. Consequently, we can get not only numerical reduction of optical response merely but also the change of dynamic transmittance from applied voltage exactly

• 한국지구물리탐사학회:학술대회논문집
• /
• 한국지구물리탐사학회 2008년도 공동학술대회
• /
• pp.51-56
• /
• 2008

등방성 매질에서의 파형역산에 대한 연구는 1980년대부터 꾸준히 이루어져 왔으나 이방성 매질에 대한 연구는 그렇지 못하다. 본 연구에서는 이방성 매질에 대한 시간영역 셀기반 유한 차분 모델링 기법을 이용해 2차원 TI 구조에서의 파형역산 알고리듬을 개발하였다. 반복적인 비선형 역산에서 최대 급경사 방향은 역시간 구조보정의 역전파 방법을 이용하여 간접적으로 계산하였고, 이를 정규화 시키기 위해 슈도-헤시안 행렬을 이용하였다. 본 연구에서 제시된 시간영역 파형역산 기법을 이방성 매질을 포함한 2층 구조와 이방성 Marmousi 모형 자료에 적용하고 이를 등방성 매질만을 고려한 기존의 파형역산 결과와 비교하였다. 본 연구의 결과를 통해 이방성 매질을 등방성 매질로 가정하고 파형역산을 수행할 경우 정확한 영상을 얻을 수 없기 때문에, 실제 탐사 자료의 파형역산을 수행할 경우 이방성 매질을 고려해야 좀 더 정확한 지하 구조를 파악할 수 있음을 확인하였다.

• 한국전기전자재료학회논문지
• /
• 제24권10호
• /
• pp.822-833
• /
• 2011

A Xe plasma flat lamp, which has been noticed as a new eco-friendly LCD (liquid crystal display) backlight, requires the improvement of the luminance and the luminous efficiency although it has several advantages. To improve the performance of a lamp, it is necessary to understand the effects of discharge variables on the luminous characteristics of the lamp. Since it is difficult to diagnose a lamp discharge experimentally, the numerical analysis can be used instead. In this study, the luminous characteristics of a planar type Xe plasma flat lamp were analyzed with the variation of an input voltage and a pulse frequency. The numerical analysis of a lamp discharge was then performed using a RCT (relaxation continuum) model and a LFA (local field approximation) model. The comparison with the experimental results showed that the RCT model is valid for the numerical analysis of the flat lamp. The numerical analysis also showed that the modifications of a high frequency component and a voltage falling rate in the input voltage waveform could improve the luminous characteristics of the lamp.

• 한국해양공학회지
• /
• 제36권1호
• /
• pp.11-20
• /
• 2022

Generally, tsunamis are generated by the rapid crustal movements of the ocean floor. Other factors of tsunami generation include landslides on coastal and ocean floor slopes, glacier collapses, and meteorite collisions. In this study, two numerical analyses were conducted to examine the formation, propagation, and deformation properties of landslide tsunamis. First, LS-DYNA was adopted to simulate the formation and propagation processes of tsunamis generated by dropping rigid bodies. The generated tsunamis had smaller wave heights and wider waveforms during their propagation, and their waveforms and flow velocities resembled those of theoretical solitary waves after a certain distance. Second, after the formation of the landslide tsunami, a tsunami based on the solitary wave approximation theory was generated in a numerical wave tank (NWT) with a computational domain that considered the stability/steady phase. The comparison of two numerical analysis results over a certain distance indicated that the waveform and flow velocity were approximately equal, and the maximum wave pressures acting on the upright wall also exhibited similar distributions. Therefore, an effective numerical model such as LS-DYNA was necessary to analyze the formation and initial deformations of the landslide tsunami, while an NWT with the wave generation method based on the solitary wave approximation theory was sufficient above a certain distance.

• 전자공학회논문지A
• /
• 제28A권1호
• /
• pp.8-14
• /
• 1991

A new algorithm for calculation of overcurrent and overvoltage at load for parallel two-wire transmission line with nonlinear protection circuits induced by and external electromagnetic pulse is suggested. The rigorous solution is obtained for a particular type of the incident waveform and protection circuit. The validity of our algorithm is checked by comparing numerical results to the analytic solution in the particular case.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2000년도 ITC-CSCC -2
• /
• pp.1096-1099
• /
• 2000

A method to analyze the high speed inter-connects that are composed of frequency dependent lossy distributed lines is presented. Network modeling of hybrid systems is implemented by using the modified nodal admittance matrix in the Laplace transformation domain. The network response is computed by different two methods. One method Is the asymptotic waveform evaluation (AWE) method and other is numerical Laplace inversion method. The merits and demerits of two methods are discussed by applying to several concrete illustrative networks.

• 한국지구과학회지
• /
• 제37권5호
• /
• pp.277-285
• /
• 2016

We introduce a depth scaling strategy to improve the accuracy of frequency-domain elastic full waveform inversion (FWI) using the new pseudo-Hessian matrix for seismic data without low-frequency components. The depth scaling strategy is based on the fact that the damping factor in the Levenberg-Marquardt method controls the energy concentration in the gradient. In other words, a large damping factor makes the Levenberg-Marquardt method similar to the steepest-descent method, by which shallow structures are mainly recovered. With a small damping factor, the Levenberg-Marquardt method becomes similar to the Gauss-Newton methods by which we can resolve deep structures as well as shallow structures. In our depth scaling strategy, a large damping factor is used in the early stage and then decreases automatically with the trend of error as the iteration goes on. With the depth scaling strategy, we can gradually move the parameter-searching region from shallow to deep parts. This flexible damping factor plays a role in retarding the model parameter update for shallow parts and mainly inverting deeper parts in the later stage of inversion. By doing so, we can improve deep parts in inversion results. The depth scaling strategy is applied to synthetic data without lowfrequency components for a modified version of the SEG/EAGE overthrust model. Numerical examples show that the flexible damping factor yields better results than the constant damping factor when reliable low-frequency components are missing.