• International journal of advanced smart convergence
• /
• 제6권4호
• /
• pp.67-72
• /
• 2017

Deep learning neural network becomes very popular nowadays due to the reason that it can learn a very complex dataset such as the image dataset. Although deep learning neural network can produce high accuracy on the image dataset, it needs a lot of time to reach the convergence stage. To solve the issue, we have proposed a scaling method to improve the neural network to achieve the convergence stage in a shorter time than the original method. From the result, we can observe that our algorithm has higher performance than the other previous work.

• 대한기계학회논문집A
• /
• 제22권2호
• /
• pp.380-390
• /
• 1998

In the present work a finite element formulation using dynamic explicit time integration scheme is used for numerical analysis of auto-body panel stamping processes. The lumping scheme is employed for the diagonal mass matrix and linearizing dynamic formulation. A contact scheme is developed by combining the skew boundary condition and direct trial-and-error method. In this work, for economic analysis the faster punch velocity and the mass scaling method are introduced. To investigate the effects of punch velocity and mass scaling, the various values of punch velocity and the various mass scalings are used for numerical analysis. Computations are carried out for analysis of complicated auto-body panel stamping processes such as forming of an oil pan and a fuel tank.

• Journal of Mechanical Science and Technology
• /
• 제15권6호
• /
• pp.764-775
• /
• 2001

Oscillartory thermocapillary flow of high Prandtl number fluids in the half-zone configuration is investigated. Based on experimental observations, one oscillation cycle consists of an active period where the surface flow is strong and the hot corner region is extended and a slow period where the opposite occurs. It is found that during oscillations the deformation of free surface plays an important role and a surface deformation parameter S correlates the experimental data well on the onset of oscillations. A scaling analysis is performed to analyze the basic steady flow in the parametric ranges of previous ground-based experiments and shows that the flow is viscous dominant and is mainly driven in the hot corner. The predicted scaling laws agree well with the numerical results. It is postulated that the oscillations are caused by a time lag between the surface and return flows. A deformation parameter S represents the response time of the return flow to the surface flow.

• Journal of Computing Science and Engineering
• /
• 제4권3호
• /
• pp.189-206
• /
• 2010

It is generally accepted that dynamic voltage scaling (DVS) is one of the most effective techniques of energy minimization for real-time applications in embedded system design. The effectiveness comes from the fact that the amount of energy consumption is quadractically proportional to the voltage applied to the processor. The penalty is the execution delay, which is linearly and inversely proportional to the voltage. According to the granularity of tasks to which voltage scaling is applied, the DVS problem is divided into two subproblems: inter-task DVS problem, in which the determination of the voltage is carried out on a task-by-task basis and the voltage assigned to the task is unchanged during the whole execution of the task, and intra-task DVS problem, in which the operating voltage of a task is dynamically adjusted according to the execution behavior to reflect the changes of the required number of cycles to finish the task before the deadline. Frequent voltage transitions may cause an adverse effect on energy minimization due to the increase of the overhead of transition time and energy. In addition, DVS needs to be carefully applied so that the dynamically varying chip temperature should not exceed a certain threshold because a drastic increase of chip temperature is highly likely to cause system function failure. This paper reviews representative works on the theoretical solutions to DVS problems regarding inter-task DVS, intra-task DVS, voltage transition, and thermal-aware DVS.

• 한국연소학회지
• /
• 제5권2호
• /
• pp.51-62
• /
• 2000

The relationship among the flame radiation, NOx emissions, residence time, and global strain rate are examined for turbulent non-premixed jet flames with wide variations in coaxial air conditions. Measurements of NOx emission, flame geometry and flame radiation were made to explain the NOx emission scaling based on global parameters such as flame residence time, global strain rate, and radiant fraction. The overall 1/2-power scaling is observed in coaxial air flames, irrespective of coaxial air conditions, but the degree of deviation from the 1/2-slope curve in each case differs from one another. From the comparison between the results of pure hydrogen flames and those of helium diluted hydrogen flames, it is observed that flame radiation plays a significant role in pure hydrogen flames with coaxial air and the deviation from 1/2-power scaling may be explained in two reasons: the difference in the flame radiation and the difference in jet similarity in coaxial air flames. From the radiation measurements, more detailed explanations on these deviations were suggested.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2018년도 학술발표회
• /
• pp.176-176
• /
• 2018

This study presents a regional, probabilistic framework for estimating streamflow via spatial scaling in the Great Lakes basin, which is the largest lake system in the world. The framework follows a two-fold strategy including (1) a quadratic-programming based optimization model a priori to explore the model structure, and (2) a time-varying hierarchical Bayesian model based on insights found in the optimization model. The proposed model is developed to explore three innovations in hierarchical modeling for reconstructing historical streamflow at ungaged sites: (1) information of physical characteristics is utilized in spatial scaling, (2) a time-varying approach is introduced based on climate information, and (3) heteroscedasticity in residual errors is considered to improve streamflow predictive distributions. The proposed model is developed and calibrated in a hierarchical Bayesian framework to pool regional information across sites and enhance regionalization skill. The model is validated in a cross-validation framework along with four simpler nested formulations and the optimization model to confirm specific hypotheses embedded in the full model structure. The nested models assume a similar hierarchical Bayesian structure to our proposed model with their own set of simplifications and omissions. Results suggest that each of three innovations improve historical out-of-sample streamflow reconstructions although these improvements vary corrsponding to each innovation. Finally, we conclude with a discussion of possible model improvements considered by additional model structure and covariates.

• 한국통신학회논문지
• /
• 제12권1호
• /
• pp.27-35
• /
• 1987

유한어장으로 인하여 디지털 필터에는 반올림한 오차의 잡음과 가산기의 오버플로 문제가 있다. 오버플로가 발생하지 않으면서 내부신호의 레벨을 가능한 최대가 되도록 바꾸기 위하여 행하는 것이 스케일링의 목적이다. 스케일링의 필요조건은 변압기를 사용하여 실현한다. 본 논문에서 오버플로를 피하면서 출력의 SNR을 극대화하는 절차를 제안하였다. 그 결과 스케일링한 경우에 오버플로가 발생하지 않으면서 SNR이 개선되었다.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2002년도 합동 추계학술대회 논문집 정보 및 제어부문
• /
• pp.282-285
• /
• 2002

In this paper, we designed of low power system by using dynamic scaling. As an effective low-power design, dynamic voltage/frequency scaling recently has received a lot of attention. In dynamic frequency scheme, all execution cycles are driven by the clock frequency that switched frequency dynamically at run time. The algorithm schedules lower frequency operators at earlier steps and higher frequency operators to later steps. This algorithm assigned the frequency for each execution cycle then it adjusted the voltage associated with the frequency.

• Journal of Astronomy and Space Sciences
• /
• 제35권3호
• /
• pp.143-149
• /
• 2018

The ionosphere has been monitored by ionosondes for over five decades since the 1960s in Korea. An ionosonde typically produces an ionogram that displays radio echoes in the frequency-range plane. The trace of echoes in the plane can be read either manually or automatically to derive useful ionospheric parameters such as foF2 (peak frequency of the F2 layer) and hmF2 (peak height of the F2 layer). Monitoring of the ionosphere should be routinely performed in a given time cadence, and thus, automatic scaling of an ionogram is generally executed to obtain ionospheric parameters. However, an auto-scaling program can generate undesirable results that significantly misrepresent the ionosphere. In order to verify the degree of misrepresentation by an auto-scaling program, we performed manual scaling of all 35,136 ionograms measured at Jeju ($33.43^{\circ}N$, $126.30^{\circ}E$) throughout 2012. We compared our manually scaled parameters (foF2 and hmF2) with auto-scaled parameters that were obtained via the ARTIST5002 program. We classified five cases in terms of the erroneous scaling performed by the program. The results of the comparison indicate that the average differences with respect to foF2 and hmF2 between the two methods approximately correspond to 0.03 MHz and 4.1 km, respectively with corresponding standard deviations of 0.12 MHz and 9.58 km. Overall, 36 % of the auto-scaled results differ from the manually scaled results by the first decimal number. Therefore, future studies should be aware of the quality of auto-scaled parameters obtained via ARTIST5002. Hence, the results of the study recommend the use of manually scaled parameters (if available) for any serious applications.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2003년도 하계종합학술대회 논문집 Ⅲ
• /
• pp.1379-1382
• /
• 2003

Recently, mobile embedded systems used widly in various applications. Managing power consumption is becoming a matter of primary concern because those systems use limited power supply. As an approach reduce power consumption, voltage can be scaled down. according to the execution time and deadline. By reducing the supplying voltage to 1/N power consumption can be reduced to 1/N. DPM-S is a well known method for dynamic voltage scaling. In this paper, we enhanced DPM-S by using average execution time aggressively. The frequency of processor is calculated based in average execution time instead of worst case execution time. Simulation results show that our method achieve up to 5％ energy savings than DPM-S.