• Structural Engineering and Mechanics
• /
• 제34권1호
• /
• pp.109-133
• /
• 2010

This paper addresses the modified Dynamic Relaxation algorithm, called mdDR by minimizing displacement error between two successive iterations. In the mdDR method, new relationships for fictitious mass and damping are presented. The results obtained from linear and nonlinear structural analysis, either by finite element or finite difference techniques; demonstrate the potential ability of the proposed scheme compared to the conventional DR algorithm. It is shown that the mdDR improves the convergence rate of Dynamic Relaxation method without any additional calculations, so that, the cost and computational time are decreased. Simplicity, high efficiency and automatic operations are the main merits of the proposed technique.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 1998년도 The Third Asian Fuzzy Systems Symposium
• /
• pp.113-117
• /
• 1998

We propose a way to show the inherent learning complexity for the multilayer perceptron. We display the solution space and the error surfaces on the input space of a single neuron with two inputs. The evolution of its weights will follow one of the two error surfaces. We observe that when we use the back-propagation(BP) learning algorithm (1), the wight cam not jump to the lower error surface due to the implicit continuity constraint on the changes of weight. The self-relaxation approach is to explicity find out the best combination of all neurons' two error surfaces. The time complexity of training a multilayer perceptron by self-relaxationis exponential to the number of neurons.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1993년도 정기총회 및 추계학술대회 논문집 학회본부
• /
• pp.228-230
• /
• 1993

In this paper, to investigate the activation energy, dipole moment, relaxation time, and so on related to charged particles more completely, it was applied numerical method of asymptotic estimation to separate single relaxation from TSC spectra with a complex relaxation. As a result, we could calculated the error of physical factors related to charged particles of specimen, more accurately.

• 대한간호학회지
• /
• 제30권2호
• /
• pp.282-292
• /
• 2000

A meta-analysis of 14 quasi-experimental studies was conducted to compare the effect of size on various relaxation therapies applied to patients and health volunteer students. These studies were selected from theses, dissertations and papers that have been done between 1982 to 1993. Also They have a randomized or nonequivalent control group in a pre test-post test design. The studies were evaluated in different ways; 1) types of relaxation therapy, 2) total amount of time of relaxation therapy, and 3) types of outcome variables. For a group of homogenious studies, the weighted mean effect size and standard error were estimated. Some findings are summarized as follows : 1. Jacobson relaxation therapy had a larger effect on systolic and diastolic blood pressures than on state anxiety. 2. For the total time of relaxation therapy, (longer than 60 minutes) had a much larger effect in decreasing systolic and diastolic blood pressures than in the case of a time period shorter than 60 minutes. 3. Relaxation therapy applied to surgery patients also had a larger effect in decreasing state anxiety than when applied to other patients.

• 대기
• /
• 제20권3호
• /
• pp.287-306
• /
• 2010

An ensemble Kalman filter (EnKF) with Weather Research and Forecasting (WRF) Model is applied for Typhoon Wukong (200610) to investigate the performance of ensemble forecasts depending on experimental configurations of the EnKF. In addition, the ensemble sensitivity analysis is applied to the forecast and analysis ensembles generated in EnKF, to investigate the possibility of using the ensemble sensitivity analysis as the adaptive observation guidance. Various experimental configurations are tested by changing model error, ensemble size, assimilation time window, covariance relaxation, and covariance localization in EnKF. First of all, experiments using different physical parameterization scheme for each ensemble member show less root mean square error compared to those using single physics for all the forecast ensemble members, which implies that considering the model error is beneficial to get better forecasts. A larger number of ensembles are also beneficial than a smaller number of ensembles. For the assimilation time window, the experiment using less frequent window shows better results than that using more frequent window, which is associated with the availability of observational data in this study. Therefore, incorporating model error, larger ensemble size, and less frequent assimilation window into the EnKF is beneficial to get better prediction of Typhoon Wukong (200610). The covariance relaxation and localization are relatively less beneficial to the forecasts compared to those factors mentioned above. The ensemble sensitivity analysis shows that the sensitive regions for adaptive observations can be determined by the sensitivity of the forecast measure of interest to the initial ensembles. In addition, the sensitivities calculated by the ensemble sensitivity analysis can be explained by dynamical relationships established among wind, temperature, and pressure.

• 한국멀티미디어학회논문지
• /
• 제6권5호
• /
• pp.788-796
• /
• 2003

본 논문에서는 순차적 학습 방법에서의 동적 모멘트를 제안한다. 동적 모멘트에서의 가변적인 모멘트를 이용하여 수렴 속도와 학습 성능을 향상시키며 회귀율에서도 이를 확인할 수 있다 제안된 학습 방법은 기존의 정적모멘트와는 달리 수렴 정도에 따라 현재의 학습에 과거의 학습률을 달리 반영하는 방법이다. 기존의 정적 상수로 정의된 모멘트가 전체 학습에 동등하게 영향을 주는 반면 제안된 동적모멘트를 이용한 학습 방법은 학습 수행에 따라 동적으로 모멘트를 변경함으로써 수렴 속도와 학습 성능을 효과적으로 제어할 수 있다. 이전의 분류문제와 회귀문제의 분리확인과는 달리 본 논문에서는 제안된 동적모멘트의 성능과 회귀율을 동시에 확인한다. 본 논문에서 사용한 회귀방법은 RMS 오류율을 사용하였으며 제안된 학습방법인 동적모멘트를 SVM(Support Vector Machine)의 순차 학습방법인 KA(Kernel Adatron)과 KR(Kernel Relaxation)에 적용하여 RMS 오류율을 확인하였다. 공정한 학습 성능 평가를 위해 신경망 분류기표준평가데이터인 SONAR 데이터를 이용하였으며 실험 결과 동적모멘트를 이용한 학습 성능과 수렴 속도 및 RMS 오류율이 정적모멘트를 이용한 학습방법보다 향상되었음을 확인하였다.

• Journal of Welding and Joining
• /
• 제17권6호
• /
• pp.84-89
• /
• 1999

Most of ferrous b.c.c weld materials may experience martensitic transformation during rapid cooling after welding. It is well known that volume expansion due to the phase transformation could influence on the relaxation of welding residual stress. To apply this effect practically, it is a prerequisite to establish a numerical model which is able to estimate the effect of phase transformation on residual stress relaxation quantitatively. For this purpose, the analysis is carried out in two regions. i.e., heating and cooling, because the variation of material properties following a phase transformation in cooling is different in comparison with the case in heating, even at the same temperature. The variation of material properties following phase transformation is considered by the adjustment of specific heat and thermal expansion coefficient, and the distribution of residual stress in analysis is compared with that of experiment by previous study. consequently, in this study, simplified numerical procedures considering phase transformation, which based on a commercial finite element package was established through comparing with the experimental data of residual stress distribution by other researcher. To consider the phase transformation effect on residual stress relaxation, the transition of mechanical and thermal property such as thermal expansion coefficient and specific heat capacity was found by try and error method in this analysis.

• 한국통신학회논문지
• /
• 제33권12C호
• /
• pp.1082-1087
• /
• 2008

MIMO 시스템에서 ML 검출 기법은 많은 다른 검출기들보다 우수한 성능을 보인다. 그러나 ML 검출기법은 NP-hard 문제로 인해 실제 시스템에서 사용하기 어려운 단점을 가지고 있다. 이것은 polynomial-time 안에 최적의 해 (optimal solution)를 찾을 수 없음을 의미한다. 본 논문에서는 ML problem을 적용한 SDR (Semi-Definite Relaxation)에 궤환기법을 통한 검출 알고리즘을 제안한다. 이는 SDR에 의해 구한 최적의 해를 spectral decomposition을 이용해 우세한 eigenvector를 찾아 송신 신호의 확률 분포를 구하고, 이를 수신 신호에 궤환 시킨다. 이는 또 다른 ML problem으로써 다시 SDR를 통해 최적의 해를 구하고 우세한 eigenvector에 해당하는 송신 신호 확률을 구한다. 이 확률은 ML problem에 해당하는 최적의 값으로 추정 송신 신호를 검출할 수 있다. 이러한 기법을 통해 최적 성능을 갖는 ML 검출 기법의 성능에 보다 더 가깝게 접근하였다.

• 소성∙가공
• /
• 제13권6호
• /
• pp.509-514
• /
• 2004

An analysis using FEM simulation was conducted to predict residual stresses and birefringence in simple compressed cylindrical glass as a preliminary part of the optimum design determination of optical lenses. The FEM simulation with the Maxwell viscoelastic constitutive model was used to predict thermal induced residual stresses and birefringence during the compression test considering stress relaxation. Also the linear photoelastic theory was introduced to calculate birefringence from the residual stress state. The error of simulation results between experimental results in the birefringence value at the center of glass specimen is $4.2\%$, and the error in the maximum radius of deformed glass specimen is $1.2\%$. The simulation results were in good agreement with deformation and birefringence distribution in the existing experimental result.

• 한국지능시스템학회논문지
• /
• 제12권2호
• /
• pp.151-156
• /
• 2002

학업성적 평가방법에는 일반적으로 절대평가와 상대평가가 있다. 이들 방법은 각기 장점들을 가지고 있으나 평가기준이 너무 일률적이거나 아니면 학습자간에 지나친 경쟁을 유발하는 등 일부 한계점들을 내포하고 있다. 본 논문에서는 일률적인 기준이 아닌 학습자들의 점수와 발생빈도를 감안한 등급화를 통해 학업성적을 평가하는 방법을 제안한다. 즉 전체 학습자의 학업성적을 적합도(goodness)를 감안하여 몇 개의 클래스로 분할하고 이를 등급으로 하는 것이다. 이를 위해 본 논문은 RE(Relaxation Error)를 적용하여 적합도를 구하고 이를 바탕으로 성적을 등급화 한다. 제안된 방법은 학업성적의 분포와 발생빈도를 고려한 적합도 측정에 의해 등급기준을 정하고 이에 따라 학업성적을 등급화 하므로 근소한 차고 등급이 구분되는 사례가 거의 없게 되며, 학교간 학생들의 학업 성적도를 차별화 할 수 있고, 실제적인 우수자나 부진아 선발 및 수준별 이동식 수업을 위한 평가에 유용하므고 기존의 두 방법을 보완하는 특성을 갖고 있다.