• Journal of applied mathematics & informatics
• /
• 제7권2호
• /
• pp.409-438
• /
• 2000

In this paper we develop a new procedure to control stepsize for Runge- Kutta methods applied to both ordinary differential equations and semi-explicit index 1 differential-algebraic equation In contrast to the standard approach, the error control mechanism presented here is based on monitoring and controlling both the local and global errors of Runge- Kutta formulas. As a result, Runge-Kutta methods with the local-global stepsize control solve differential of differential-algebraic equations with any prescribe accuracy (up to round-off errors)

• 한국통신학회논문지
• /
• 제27권5C호
• /
• pp.474-485
• /
• 2002

FIFO 기반 버퍼관리 알고리즘의 공평성 향상과 단순성을 유지하기 위해 새로운 LBO threshold 결정 메커니즘을 제안하였다. 제안한 알고리즘은 활성화된 각 VC의 가중치 합에 따라 LBO를 차별적으로 계산하여 동적으로 적용한다. 이러한 LBO 설정 메커니즘을 DFBA 알고리즘에 적용한 DT-DFBA(Differential Threshold-Differential Fair Buffer Allocation)을 구현하고 실험을 통하여 성능을 평가하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제12권7호
• /
• pp.3497-3515
• /
• 2018

Differential privacy has broadly applied to statistical analysis, and its mainly objective is to ensure the tradeoff between the utility of noise data and the privacy preserving of individual's sensitive information. However, an individual could not achieve expected data utility under differential privacy mechanisms, since the adding noise is random. To this end, we proposed an adaptive Gaussian mechanism based on expected data utility under conditional filtering noise. Firstly, this paper made conditional filtering for Gaussian mechanism noise. Secondly, we defined the expected data utility according to the absolute value of relative error. Finally, we presented an adaptive Gaussian mechanism by combining expected data utility with conditional filtering noise. Through comparative analysis, the adaptive Gaussian mechanism satisfies differential privacy and achieves expected data utility for giving any privacy budget. Furthermore, our scheme is easy extend to engineering implementation.

• 정보처리학회논문지C
• /
• 제13C권3호
• /
• pp.323-330
• /
• 2006

본 논문은 차등 RSSI(DRSSI: Differential Received Signal Strength Indicator)를 이용하여 자원을 효율적으로 사용할 수 있는 핸드오프 처리 방안에 대해 제안한다. DRSSI는 이동국의 이동 방향을 예측하는데 사용할 수 있다. 즉, 이동국의 움직임이 바뀌는 경우 DRSSI의 극성(+, -)이 바뀌는 특성을 이용하여 이동국이 셀 사이의 중첩 지역에 머무는 동안 극성의 변화를 우선적으로 판단하여 다수의 단말기의 방향을 예측함으로써 핸드오프 및 자원을 예약할 이동국을 예측하고, 자원의 우선적인 사용권을 부여하는 방안을 제시하여 자원의 유효 상태를 극대화하는 방안을 제시한다. 또한, 수학적 모델링 분석을 통해 핸드오프 차단 확률이 기존의 방법보다 우수함을 보여준다.

• Journal of applied mathematics & informatics
• /
• 제6권3호
• /
• pp.697-726
• /
• 1999

In this paper we develop a now procedure to control stepsize for linear multistep methods applied to semi-explicit index 1 differential-algebraic equations. in contrast to the standard approach the error control mechanism presented here is based on monitoring and contolling both the local and global errors of multistep formulas. As a result such methods with the local-global stepsize control solve differential-algebraic equation with any prescribed accuracy (up to round-off errors). For implicit multistep methods we give the minimum number of both full and modified Newton iterations allowing the iterative approxima-tions to be correctly used in the procedure of the local-global stepsize control. We also discuss validity of simple iterations for high accuracy solving differential-algebraic equations. Numerical tests support the the-oretical results of the paper.

• 대한기계학회논문집B
• /
• 제27권7호
• /
• pp.938-947
• /
• 2003

In this paper, we derived the formula for estimating the power of the electric motors needed to operate the Decanter-type centrifuge. In the derivation of the formula the sludge-removal torque is to be supplied from the formula derived in the first paper. The intricate nature of the transmission mechanism in the planetary gear trains of the sludge-removal power and torque has been clarified in this second paper. In particular we considered two-motor system, where the main motor drives the machine while the differential-speed control motor plays the role of braking in adjusting the differential speed. Sample calculation for the specific design treated in the first paper showed that the selection criterion for the main motor depends on the lower limit of the differential speed; when the lower limit is set low, it should be selected based on the steadily operating power, while it should be selected based on the starting power when the lower limit is set high. The total power required by both the main motor and the differential-speed control motor increases as the differential speed is decreased. It is suggested that the power loss in the differential-speed control motor could be minimized by attaching an electric generator to it.

• International Journal of Precision Engineering and Manufacturing
• /
• 제2권1호
• /
• pp.43-55
• /
• 2001

A continuously variable transmission(CVT) mechanism composed of one differential gear unit and one continuously variable unit(CVU) can be classified according to the coupling of CVU and the direction of power flows. The mechanism has many advantages which are the decrease of CVT size, the increase of overall efficiency, the extension of speed ratio range and generation of geared neutral. The CVT mechanism considered here is the input coupled type which combines the functions of a 2K-H I type differential gear unit and a V-belt type CVU. One shaft of the CVU is connected directly to the input shaft and another shaft of it is linked to the differential gear unit. It is shown that some fundamental relations(speed ratios, power flows and efficiencies) for twelve mechanisms previously described are valid by various experimental studies, six of them produce a power circulation and the others produce a power split. Some useful comparisons between theoretical analysis and experimental results are presented. General properties also are discussed, which connect following power flow modes : (a) power circulation mode; (b) power split mode.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2003년도 ICCAS
• /
• pp.859-864
• /
• 2003

This paper presents a design for the novel suspension mechanism of a two-wheeled mobile robot having two casters which is used for indoor environment. Although the indoor environment is less rough than the outdoor one, the fixed caster mechanism has some problems such as causing the robot to be immovable because robot's driving wheels do not have contact with the ground. Therefore, we tried installing a spring-damper suspension mechanism to keep driving capability and to remove pitching phenomenon. However, this suspension mechanism also has the problem, which the robot body inclined by disturbances does not return to the initial position. To deal with above problems and to accomplish desired performances, we designed the Multilayered Suspension Mechanism, which has springs and dampers working partially according to the inclined angle and angular velocity of robot body concerned with pitching. To analyze design, the equations of motion describing their dynamics were developed. Using the equations, simulation results show the improved performance. We confirm the usefulness of the Multilayered Suspension Mechanism by construction and test of a actual prototype.

• 대한기계학회논문집
• /
• 제14권3호
• /
• pp.526-533
• /
• 1990

본 연구에서는 단일 캠 구동기구의 안정성 해석을 위한 효율적인 방법이 제시 되고, 수치실험을 통하여 검증되었다. 우선 단일 캠 구동기구를 묘사하는 비선형 방 정식이 유도되고, 동안정성에 관여하는 변수들이 무차원화되어 도입된다. 유도된 방 정식은 캠 축의 입력각에 대하여 선형화되고, 선형화된 방정식에 조화균형법(harmonic balance method)을 적용하여 동안정성 해석을 수행하였다.

• 정보보호학회논문지
• /
• 제34권1호
• /
• pp.143-156
• /
• 2024

연합 학습(FL)은 여러 공동 작업자 간에 분산된 모델 학습을 위한 강력한 방법론으로 부상해 데이터 공유의 필요성을 없애준다. FL은 데이터 프라이버시를 보호하는 기능으로 호평을 받고 있지만, 다양한 유형의 프라이버시 공격으로부터 자유롭지 않다. 대표적인 개인정보 보호 기술인 차분 프라이버시(DP)는 이러한 취약점에 대응하기 위해 널리 사용된다. 이 논문에서는 기존의 작업별 적응형 DP 메커니즘을 FL 환경에 적용해 성능을 평가한다. 포괄적인 분석을 통해 다양한 DP 메커니즘이 공유 글로벌 모델의 성능에 미치는 영향을 평가하며, 특히 다양한 데이터 배포 및 분할 스키마에 주의를 기울인다. 이를 통해, FL에서 개인정보 보호와 유용성 간의 복잡한 상호 작용에 대한 이해를 심화하고, 성능 저하 없이 데이터를 보호할 수 있는 검증된 방법론을 제공한다.