• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.32 no.9
• /
• pp.1-11
• /
• 2004

A comprehensive study has been made for the investigation of the convergence and stability characteristics of the LU scheme for solving the Euler equations on unstructured meshes. The von Neumann stability analysis technique was initially applied to a scalar model equation, and then the analysis was extended to the Euler equations. The results indicated that the convergence rate is governed by a specific combination of flow parameters. Based on this insight, it was shown that the LU scheme does not suffer any convergence deterioration at all grid aspect ratios, as long as the local time step is defined using an appropriate parameter combination.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.23 no.7
• /
• pp.1869-1882
• /
• 1998

In this paper, we investigae convergence behaviors of a differential method for 2-D motion parameter estimation. While the differential method is widely studied for motion compensated prediction in video coding, little attention hs been paid to its convergence properties. Based on the nonseparable exponential covariance image model, we derive the estimates of update terms for the 2- and 6-parameter motion models. And, the effect of noise, spatial correlation, choice of spatial gradient measures, andthe size of a region, are quantitatively anlyzed in relation to the convergence speed. Some empirical results are presented to verify the analysis.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.32 no.7C
• /
• pp.627-634
• /
• 2007

When a higher-order modulation scheme (16QAM or 64QAM) is applied to the communications system using the nonlinear high power amplifier (HPA), the performance can be degraded by the nonlinear distortion of the HPA. The nonlinear distortion can be compensated by the adaptive nonlinear Volterra equalizer using the low-complexity LMS algorithm at the receiver. However, the LMS algorithm shows very slow convergence performance. So, in this paper, the parallel M-band discrete wavelet transformed LMS algorithm is proposed in order to improve the convergence speed. Throughout the computer simulations, it is shown that the convergence performance of the proposed method is superior to that of the conventional time-domain and transform-domain LMS algorithms.

• Computational Structural Engineering
• /
• v.2 no.3
• /
• pp.97-103
• /
• 1989

In this study, a usefulness of the iterative solution procedures is reviewed for the elasto-plastic large deflection analysis of imperfect plates by finite element method. Three typical solution techniques such as simple incremental(SI) method, Newton-Raphson(NR) method and modified Newton-Raphson (mNR) method are compared. It is concluded that for thin plates which are given rise to the large deflection, iteration for the convergence of the unbalance force should be performed and in this case mNR method is more useful than NR method since the computing time of the former becomes to be a half of the latter, in which the accuracy of the result remains same. For thick plates or thin plates with large initial deflection, however, the use of SI method is quite better since the unbalance force may be negligible.

• The Journal of the Korea Contents Association
• /
• v.6 no.5
• /
• pp.29-34
• /
• 2006

For the almost certainly convergent series $S_n=\sum_{i=1}^nV-i$ of independent random elements in Banach spaces, by investigating tail series laws of large numbers, the rate of convergence of the series $S_n$ to a random variable s is studied in this paper. More specifically, by studying the duality between the limiting behavior of the tail series $T_n=S-S_{n-1}=\sum_{i=n}^{\infty}V-i$ of random variables and that of Banach space valued random elements, an alternative way of proving a result of the previous work, which establishes the equivalence between the tail series weak law of large numbers and a limit law, is provided in a Banach space setting.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.8 no.10
• /
• pp.1451-1456
• /
• 2013

This paper propose a method for fast convergence technique in frequency domain independent component analysis (FDICA) using eigenanalysis. It important, such as SONAR system, to eliminate the interference sources through fast algorithm. Through eigenanalysing a two-by-two delayed mixture case, information of delay can be used for initial weighting parameters. Simulations show the improved performances in convergence speed and noise rejection rate. The proposed method can present close weights for optimal convergence, noise can be diminished drastically about 3 times epoch, and get the better resultss with 1~3dB than the conventional method.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2015.07a
• /
• pp.457-459
• /
• 2015

본 논문에서는 원의 호 곡선에 따라 배치된 다중 RGB 카메라 영상으로 생성한 깊이 지도를 정렬하는 방법을 제안한다. 원의 호 곡선에 따라 배치된 카메라는 각 카메라의 광축이 한 점으로 만나서 수렴하는 형태가 이상적이다. 그러나 카메라 파라미터를 살펴보면 광축이 서로 수렴하지 않는다. 또한 카메라 파라미터는 오차가 존재하고 내부 파라미터도 서로 다르기 때문에 각 카메라 영상들은 수평과 수직 오차가 발생한다. 이와 같은 문제점을 해결하기 위해 첫 번째로 광축이 한 점으로 수렴하기 위해서 카메라 외부 파라미터를 보정하여 깊이 영상 정렬을 하였다. 두 번째로 내부 파라미터를 수정하여 각 깊이 영상들의 수평과 수직 오차를 감소시켰다. 일반적으로 정렬된 깊이 지도를 얻기 위해서는 초기 RGB 카메라 영상으로 정렬을 수행하고 그 결과 영상으로 깊이 영상을 생성한다. 하지만 RGB 영상으로 카메라의 회전과 위치를 보정하여 정렬하면 카메라 위치 변화에 따른 깊이 지도 변화값 적용이 복잡해 진다. 즉 정렬 계산 과정에서 소수점 단위 값이 사라지기에 최종 깊이 지도의 값에 영향을 미친다. 그래서 RGB 영상으로 깊이 지도를 생성하고 그것을 처음 RGB 카메라 파라미터로 워핑(warping)하였다. 그리고 워핑된 깊이 지도 값을 가지고 정렬을 수행하였다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• v.9 no.1
• /
• pp.447-452
• /
• 2005

기존의 단층 퍼셉트론은 출력 노드가 선형 분리 가능한 패턴들만을 분류할 수 있고 Exclusive OR와 같은 비선형 문제에 대해서는 분류할 수 없는 단점이 있다. 그러나 퍼지 단층 퍼셉트론은 퍼지소속 함수(fuzzy membership function)를 적용하여 단층 구조로 Exclusive OR 문제와 같은 고전적인 문제를 개선하였다. 그러나 퍼지 단층 퍼셉트론은 기존의 단층 퍼셉트론과 마찬가지로 결정 경계선이 진동하는 경우가 생기며 초기 가중치의 범위와 학습률에 따라 수렴성이 매우 낮아지는 단점이 있다. 따라서 본 논문에서는 바이어스항을 도입하여 결정 경계선이 진동하는 것을 방지하여 수렴성을 개선시키고 선형 활성화 함수를 제안하고 학습률과 모멘텀 개념을 도입하여 학습 시간을 단축시키는 개선된 퍼지 단층 퍼셉트론 알고리즘을 제안한다. 제안된 방법과 퍼지 단층 퍼셉트론간의 학습 성능을 분석하기 위하여 인공 신경망에서 벤치마크로 사용되는 exclusive OR 문제와 문자 패턴 분류에 적용하여 epoch 수와 수렴성을 비교한 결과, 제안된 방법이 기존의 퍼지 단층 퍼셉트론보다 학습 시간이 적게 소요되고 수렴성이 개선된 것을 확인하였다.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.12 no.9
• /
• pp.4147-4152
• /
• 2011

Compared to the constant modulus algorithm (CMA), the modified CMA (MCMA) is easy not only to improve the steady-state performance but also to be expanded to higher-order constellations by using fewer moduli with evenly spaced. In this paper, it is shown that the MCMA is sufficient to achieve satisfactory steady-state performance by applying a variable step-size to the MCMA without switching to an hard decision-directed algorithm. Two new methods varying the step-size are proposed, and the performance improvement of the MCMA with the new methods of variable step-size is presented as compared to the CMA and the fixed step-size MCMA through computer simulations.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2009.10a
• /
• pp.474-476
• /
• 2009

In the radio communication, such as echo cancellation and channel equalization, adaptive filtering is very practical. Its convergence behavior that is used for updating the weights depends on the correlation of the input signal and length of adaptive filter. Highly correlated input and long length of adaptive filter deteriorate the convergence behavior. To solve this problem, recently, subband affine projection algorithm which pre-whiten the correlation of the input and update the weights in subband structure has been presented. This paper presents convergence analysis method of multiple constrained subband affine projection algorithm.