• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.5C
• /
• pp.692-699
• /
• 2004

In this paper we studied an implementation of QR decomposition-based RLS algorithm using modified Givens rotation method. Givens rotation can be obtained with a sequence of the CORDIC operations. In order to reduce the computing time of QR decomposition we restricted the number of iterations of the CORDIC operation per a Givens rotation and used double-rotation method to remove the square-root in the scaling factor.

• Proceedings of the KIEE Conference
• /
• 1999.11b
• /
• pp.248-250
• /
• 1999

본 논문에서 제안된 조류계산 알고리즘은 Jacobian의 off-diagonal 부 행렬을 완전히 무시하는 Decoupled Load Flow(DCL) 알고리즘들의 유연한 대안으로 반복(iteration)당 최소한의 추가 계산 부담으로 Jacobian의 off-diagonal 부분의 효과를 부분적으로 반영함에 의해 수렴 특성을 향상시킬 수 있다. 제안된 방법들은 특히 Fast Decoupled Load Flow(FDL)로 대표되는 DCL들의 수렴특성이 불안정해질 경우, 효과적으로 수렴특성을 향상시킬 수 있다. 제안된 알고리즘들은 Newton-Raphson Load Flow(NRL) 방법에서 Jacobian의 off-diagonal 부분의 효과를 점진적으로 제거하는 방법으로 유도하였고, 간략화 과정에서는 Neuman series expansion을 사용하였다. 실험결과 제안된 알고리즘들은 반복횟수와 전반적인 수렴 속도에서 확실한 성능향상을 보여주었다. 제안된 알고리즘들은 특히 DCL의 수렴성능이 문제가 있을 시 full NRL 대신에 적용할 수 있는 가능성이 있어 조류계산 시간을 단축해 줄 수 있을 것으로 기대된다.

• Journal of the Institute of Convergence Signal Processing
• /
• v.2 no.2
• /
• pp.13-19
• /
• 2001

The conventional fractal decoding was required a vast amount computational complexity, since every range blocks was implemented to IFS(iterated function system). In order to improve this, it has been suggested that each range block was classified to iterated and non-iterated regions. Non-iterated regions is called data dependency region, and if data dependency region extended, IFS regions are contractive. In this paper, a searched region of the domain is limited to the range regions that is similar with the domain blocks, and the domain region is more overlapped. As a result, data dependency region has maximum region, that is IFS regions can be minimum region. The minimizing method of domain region is defined to minimum domain(MD) which is minimum IFS region. Using the minimizing method of domain region, there is not influence PSNR(peak signal-to-noise ratio). And it can be performed a fast decoding by reducing the computational complexity for IFS in fractal image decoding.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.22 no.8
• /
• pp.1803-1813
• /
• 1997

The goal of image restoration is to remove the degradations in a way that the resrored image will best approximate the original image. This can be done by the iterative regularized image restoration method. In any iterative image restoration algorithm, using a "better" termination rule results in both "better" quality of ther restored image and "less" computation, and hence, "faster" and "simp;er" practical system. Therefore, finding a better termmination rule for an iterative image restoration algorithm has been an interesting and improtant question for many researchers in the iterative image restoration. In these reasons, the new termination rule using the estimated distance between the original image and the restored image is proposed inthis paper. Noise suppression parameter(NSP) and the rule for estimating NSP with the noise variance are also proposed. The experimental results shows that the proposed termination rule is superior to the conventional methods.

• Proceedings of the Korean Information Science Society Conference
• /
• 2006.06c
• /
• pp.133-135
• /
• 2006

실시간 프로그램은 다양한 분야에서 사용되고 있으며, 그 개발자는 논리적, 시간적 정확성을 고려해야 한다. 시간적 정확성은 실시간 프로그램에서 가장 중요한 부분이며, 이를 위한 데드라인은 개발자에 의해 정의된다. 따라서 개발자는 데드라인의 정의를 위하여 기준점을 제시할 수 있는 정적 실행시간 분석이 필요하다. 정적 실행시간 분석에서 프로그램의 반복횟수 분석은 큰 비중을 차지한다. 현재 이러한 반복횟수 분석을 자동화하는 연구가 진행 중이다. 하지만 반복횟수 분석은 반복횟수에 영향을 주는 제어변수의 결정정책에 따라 결과가 달라지고, 자동화를 위한 반복횟수 계산 수식의 적용 범위가 제한되어 있다. 본 논문에서는 이러한 제어변수를 결정 및 탐색하고, 수집된 정보를 개선된 수식을 사용하여 반복횟수 분석을 수행할 수 있도록 PS-Block 구조를 기반으로 반복횟수 분석기를 설계 및 구현하였다. 반복횟수 분석기는 제어변수의 탐색 및 결정 분석 과정을 자동화하고, 수식의 개선으로 자동화 범위를 확대하며, 개별 반복 문 단위의 정밀한 반복횟수 분석을 통해 정확도를 높이고, 신뢰성을 향상시킬 수 있다.

• Proceedings of the IEEK Conference
• /
• 2001.09a
• /
• pp.469-472
• /
• 2001

MPEC Audio Layer-III(MP3)알고리듬은 복호화기에 비해 부호화기가 월등히 많은 연산량을 가지고 있는 비대칭 구조를 가지고 있다. MP3 부호화기의 대부분의 연산량은 복잡한 초월함수 연산이 포함되는 심리음향모델과 반복 루프 과정을 수행하는 비선형 양자화와 비트 할당과정 이 차지한다. 본 논문에서는 MP3 부호화기의 실시간 구현을 위한 알고리듬 레벨의 최적화를 수행하였다. MP3 부호화기의 연산량을 줄이기 위해 심리음향모델을 간략화하고 반복 루프의 회수를 최소화할 수 있는 방법을 제안하였다. 프레임당 한 그래뉼의 심리음향모델 정보를 계산하여 한 프레임 내에서의 심리음향모델 정보를 추정함으로써 연산량을 45% 이상 감소시켰다. 또한 외부 반복 루프의 반복 회수를 줄이기 위하여 외부 반복 루프의 반복에 따른 스케일 팩터(Scale Factor) 및 양자화 스탭의 증가 패턴을 관찰하고 최적화된 스캐일 팩터 증가 방법을 제안하였다. 제안된 고속화 방법은 주관적 음질 평가를 통해 성능을 검증하였다.

• Journal of Elementary Mathematics Education in Korea
• /
• v.16 no.1
• /
• pp.125-143
• /
• 2012

This study is designed to see the characteristics of elementary students' arithmetic error patterns and error rates by going up grades and to draw some implications for effective instruction. For this, 580 elementary students of grade 3-6 are tested with the same subtraction, multiplication and division problems. Their errors are analyzed by the frame of arithmetic error types this study sets. As a result of analysis, it turns out that the children's performance in arithmetic get well as their grades go up and the first learning year of any kind of arithmetic procedures has the largest improvement in arithmetic performance. It is concluded that some arithmetic errors need teachers' caution, but we fortunately find that children's errors are not so seriously systematic and sticky that they can be easily corrected by proper intervention. Finally, several instructional strategies for arithmetic procedures are suggested.

• The Journal of Engineering Geology
• /
• v.31 no.2
• /
• pp.149-163
• /
• 2021

Damage characteristics of granite, marble and sandstone whose properties were different were investigated by uniaxial compression test and cyclic loading-unloading test. Strength, elastic constants and damage threshold stresses were measured by uniaxial compression test and were compared with those measured by cyclic loading-unloading test. Average rock strengths measured by cyclic loading-unloading test were either lower than or similar with those measured by uniaxial compression test. Rocks with high strength and low porosity were more sensitive to fatigue than that with low strength and high porosity. Although permanent strains caused by cyclic loading-unloading were different according to rock types, they could be good indicators representing damage characteristics of rock. Damage threshold stress of granite and marble might be measured from stress-permanent strain curves. Acoustic emissions were measured during both tests and felicity ratios which represented damage characteristics of rocks were calculated. Felicity ratio of sandstone which was weak in strength and highly porous could not be calculated because of very few measurements of acoustic emissions. On the other hand, damage threshold could be predicted from felicity ratios of granite and marble which were brittle and low in porosity. The deformation behaviors and damage characteristics of rock mass could be investigated if additional tests for various rock types were performed.

• Journal of the korean Society of Automotive Engineers
• /
• v.10 no.5
• /
• pp.33-41
• /
• 1988

내연기관에서 crankshaft는 piston에 작용하는 수직 gas force 및 관성력을 회전력으로 바꾸어 주는 역할을 한다. 그러므로 engine작동중 높은 부하를 반복적으로 받기 때문에 이에 견딜 충분한 stiffness를 유지해야 한다. Engine의 stroke 운동시에 gas force와 관성력이 변하고 crankshaft의 형상이 복잡하므로 실제적으로 응력계산은 간단하지 않다. 여기에서는 crankshaft에 작용하는 하중과 반력을 계산하고 crankshaft의 stress 및 safety factor를 계산하여 crankshaft를 설계하는 방법을 설명하고자 한다. 근래 FEM에 의한 stress analysis를 많이 수행하나, loading condition을 주기 위해서는 classical method에 의해서 구할 수 있다.

• The Journal of the Acoustical Society of Korea
• /
• v.31 no.7
• /
• pp.427-434
• /
• 2012

The overlap-add technique based on waveform similarity (WSOLA) method is known to be an efficient high-quality algorithm for time scaling of speech signal. The computational load of WSOLA is concentrated on the repeated normalized cross-correlation (NCC) calculation to evaluate the similarity between two signal waveforms. To reduce the computational complexity of WSOLA, this paper proposes a fast NCC computation method, in which NCC is obtained through pre-calculated sum tables to eliminate redundancy of repeated NCC calculations in the adjacent regions. While the denominator part of NCC has much redundancy irrespective of the time-scale factor, the numerator part of NCC has less redundancy and the amount of redundancy is dependent on both the time-scale factor and optimal shift value, thereby requiring more sophisticated algorithm for fast computation. The simulation results show that the proposed method reduces about 40%, 47% and 52% of the WSOLA execution time for the time-scale compression, 2 and 3 times time-scale expansions, respectively, while maintaining exactly the same speech quality of the conventional WSOLA.