• 전자공학회논문지C
• /
• 제36C권4호
• /
• pp.17-29
• /
• 1999

본 논문은 데이터 의존적인 CMOS 회로(예: DSP) 의 전력량을 감축하기 위한 상위 수준 합성 기법에 대한 연구이다. 상위수준 합성은 스케줄링, 자원 및 레지스터 할당의 세가지로 나우어서 수행한다. 스케줄링시의 저전력 설계의 목적은 자원할당 시 입력을 재 사용할 수 있는 가능성을 증가시키는 것이다. 스케줄링 후에 자원 및 레지스터 할당 문제는 가중차기 부가된 앙립 그래프로 표현하여 최소비용흐름 알고리즘을 수행함으로써 스위칭 동작횟수가 적은 해를 얻는다. 제안된 알고리즘은 저전력 레지스터 및 자원 할당 문제에 대하여 O({{{{ { n}^{3 } }}}}) (n은 그래프의 노드수) 시간에 최적해를 제공한다. 벤치마크 회로에 대한 실험 결과는 15%의 전력 감축 효과를 나타낸다.

• 대한교통학회:학술대회논문집
• /
• 대한교통학회 1998년도 Proceedings 제34회 추계 학술발표회
• /
• pp.236-244
• /
• 1998

When the capacity and traffic operation at signalized intersections are analyzed in Korea, the unprotected left-turn saturation flow rate, which is an important parameter for the analysis, is estimated form the USHCM model. thus, exact analysis of the left-turn is not possible because of the difference of traffic environments between two contries. In order to improve this problem, it is undertaken in this study to develop techniques for the estimation of unprotected left-turn saturation flows based on Korean drivers' data. As study intersections, signalized or unsignalized intersections on the 6, 4 and 2 lane streets are selected. the data for the saturation flow measurement and gap-acceptance behavior analysis are inputed in a notebook computer on the sites. The critical acceptance gaps of the 6, 4, and 2 lane streets are analyzed to be 6.0 secs, 4.6 secs, and 4.3 secs respectively. the average minimum headway of the left-turn vehicle was observed to be 2.6 secs. As the model to estimate unportected left-turn saturation flows, the drew model is recommended for 6 and 4 lane streets, and a graph is suggested for the 2-lane street. As the values of the parameters of the Drew model, the 2.6 secs of this study is recommended for the average minimum headway of the left-turn. But, the critical acceptance gap varies according to the approach speed of opposing traffic and driver population, it requires field survey to measure the gap of an intersection; however, the values of the gaps studied in this study may be used for the general intersections in urban area in Korean.

• 정보보호학회논문지
• /
• 제33권3호
• /
• pp.427-435
• /
• 2023

카디널리티 추정은 실생활의 많은 곳에서 사용되며, 큰 범위의 데이터를 처리하는 데 근본적 문제이다. 인터넷이 빅데이터의 시대로 넘어가며 데이터의 크기는 점점 커지고 있지만, 작은 온칩 캐시 메모리만을 이용하여 카디널리티 추정이 이뤄진다. 메모리를 효율적으로 사용하기 위해서, 지금까지 많은 방법이 제안되었다. 그러나, 이러한 알고리즘에서는 estimator 간의 노이즈 발생으로 인해 정확도가 떨어지는 일이 발생한다. 이 논문에서는 노이즈를 최소화하는데 중점을 뒀다. 우리는 여러 개의 데이터 구조를 제안하여 각 estimator가 데이터 구조 수만큼의 추정값을 가지고, 이 중 가장 작은 값을 선택하여 노이즈를 최소화한다. 실험을 통해 이 방법이 이전의 가장 좋은 방법과 비교했을 때, 플로우당 1 bit와 같은 작은 메모리를 사용하면서 더 좋은 성능을 보이는 것을 확인했다.

• International Journal of Control, Automation, and Systems
• /
• 제3권1호
• /
• pp.56-69
• /
• 2005

In this paper, it is clearly shown that the two well-known necessary and sufficient conditions mp n as generic static output feedback pole-assignment and mp + d(m+p) n+d as generic minimum d-th order dynamic output feedback pole-assignment on complex field, unbelievably, do not match up each other in strictly proper linear systems. For the analysis, a diagram analysis is newly created (which is defined by the analysis of 'convoluted rectangular/dot diagrams' constructed via node-branch conversion of the signal flow graphs of output feedback gain loops). Under this diagram analysis, it is proved that the minimum d-th order dynamic output feedback compensator for pole-assignment in m-input, p-output, n-th order systems is quantitatively decomposed into static output feedback compensator and its associated d number of arbitrary 1st order dynamic elements in augmented (m+d)-input, (p+d)-output, (n+d)-th order systems. Total configuration of the mismatched data is presented in a Table.

• KSBB Journal
• /
• 제4권2호
• /
• pp.177-184
• /
• 1989

항생물질 발효액의 균사체 분리에 있어서 Modified regenerated cellulose 막을 이용한 Cross-flow filtration으로 높은 Flux를 얻을 수 있었다. 막분리시 가급적 높은 유속, 높은 온도, 특정 균사체농도(약 20% PMV)에서 높은 Flux를 보여주었으며 Fouling 정도는 균사체 농도가 높을수록 또한 유속이 낮을수록 심각한 것으로 나타났다. 발효액을 농축 분리할 때의 Flux profile은 1단계 Sublayer형성에 의한 급속감소, 2단계 평형, 3단계 Fouling에 의한 급속감소의 세 구간으로 나눌 수 있었으며 Diafiltration의 적용으로 3단계의 Flux 감소 현상을 현저히 완화시킬 수 있었다. 한편 Diafiltration에 의한 항생물질 회수시 초기부터Diafiltration하는 것보다는 특정 PMV까지 농축시킨 후 Diafiltration하는 경우에 작업시간과 여액의 Volume을 최소화 할 수 있었으며 본 실험에서는 98%의 회수율을 얻기 위해서 Di-afiltration을 시작하는 Optimum PMV는 20%였다.

• Journal of Advanced Marine Engineering and Technology
• /
• 제12권1호
• /
• pp.53-61
• /
• 1988

Most of the recent for the flow field hydrofoil in partially cavitating condition are the ones which are linearized, and the problem of cavity end region for hydrofoil is not verified. This paper deals with a study on characteristics of end region flow field for partially cavitating hydrofoil by using a characteristics of shear turbulence flow and nonlinear cavity flow theory. The results obtained as follows : 1) Shear layer thickness is decreased gradually going to the end section of hydrofoil. When attack angle is large, it is appeared largely at the region of partial cavitation after its collapsing. 2) The fluctuation velocity of a second-degree relative direction have minimum value at the front of hydrofoil or at the end of hydrofoil. The difference for the validity of attack angle is appeared largely at the surrounding of .chi.$_{e}$ point. 3) The fluctuation velocity of transverse direction decrease from the maximum thickness of cavitation to the end of hydrofoil, but it undergoes largely the effect of pressure recovery. The difference is larger at the region of partial cavitation after its collapsing than at the of hydrofoil. 4) The distribution of Reynolds stress have maximum values at the region of partial cavitation after its collapsing and the end of foil, and the larger attack angle, the larger the distribution of value.e.

• 한국정보처리학회논문지
• /
• 제3권7호
• /
• pp.1707-1718
• /
• 1996

N개의 signal net와 소스 노드 s, 싱크 노드가 있을 때, 최소 비용 직각선분 거리 유지 트리의 s와 싱크 간의 노선 길이는 s와 싱크 간의 직각선분의 길이와 일치한다. 최소 비용 직각선분 Steiner 거리유지 트리는 최단 s와 싱크 간의 직각선분 길이를 유지하면서 전체 트리 비용이 최소가 되도록 하는 것이다. 최소 비용 직각선분 Steiner 거리 유지 트리를 찾는 휴리스틱 알고리즘이 최근에 발표된 바 있는데, 본 논문에서는 O(n2 log m)2) 계산시간 내에 비용을 최소화하는 알고리즘을 제안한다. 네트워크 흐름 이론과 최소 비용 흐름 알고리즘을 이용하여 최적화 해를 발견하였다. 더욱 일반화된 응용 연구가 제안 되었고 관련 미해결 문제를 제시하였다.

• Nuclear Engineering and Technology
• /
• 제50권1호
• /
• pp.126-131
• /
• 2018

As it is known, uranium enrichment is carried out on industrial scale by means of multistage separation facilities, i.e., separation cascades in which gas centrifuges (GCs) are connected in series and parallel. Design and construction of these facilities require significant investment. So, the problem of calculation and optimization of cascade working parameters is still relevant today. At the same time, in many cases, the minimum unit cost of a product is related to the cascade having the smallest possible number of separation elements/GCs. Also, in theoretical studies, it is often acceptable to apply as an efficiency criterion the minimum total flow to supply cascade stages instead of the abovementioned minimum unit cost or the number of separation elements. In this article, cascades with working parameter of a single GC changing from stage to stage are optimized by two of the abovementioned performance criteria and are compared. The results obtained allow us to make a conclusion about their differences.

• 대한기계학회논문집A
• /
• 제28권10호
• /
• pp.1566-1573
• /
• 2004

When a projectile travels at high speed underwater, supercavitating flow arises, in which a huge cavity is generated behind the projectile so that only the nose, i.e., the cavitator, of the projectile is wetted, while the rest of it should be surrounded by the cavity. In that case, the projectile can achieve very high speed due to the reduced drag. Furthermore if the nose of the body is shaped properly, the attendant pressure drag can be maintained at a very low value, so that the overall drag is also reduced dramatically. In this study, shape optimization technique is employed to determine the optimum cavitator shape for minimum drag, given certain operating conditions. Shape optimization technique is also used to solve the potential flow problem fur any given cavitator, which is a free boundary value problem having the cavity shape as unknown a priori. Analytical sensitivities are derived for various shape parameters in order to implement a gradient-based optimization algorithm. Simultaneous optimization technique is proposed for efficient cavitator shape optimization, in which the cavity and cavitator shape are determined in a single optimization routine.

• 대한기계학회:학술대회논문집
• /
• 대한기계학회 2003년도 추계학술대회
• /
• pp.1876-1881
• /
• 2003

When a projectile travels at high speed underwater, supercavitating flow arises, in which a huge cavity is generated behind the projectile so that only the nose, i.e., the cavitator, of the projectile is wetted, while the rest of it should be surrounded by the cavity. In that case, the projectile can achieve very high speed due to the reduced drag. Furthermore if the nose of the body is shaped properly, the attendant pressure drag can be maintained at a very low value, so that the overall drag is also reduced dramatically. In this study, shape optimization technique is employed to determine the optimum cavitator shape for minimum drag, given certain operating conditions. Shape optimization technique is also used to solve the potential flow problem for any given cavitator, which is a free boundary value problem having the cavity shape as unknown a priori. Analytical sensitivities are derived for various shape parameters in order to implement a gradient-based optimization algorithm. Simultaneous optimization technique is proposed for efficient cavitator shape optimization, in which the cavity and cavitator shape are determined in a single optimization routine.