• 대한기계학회논문집B
• /
• 제30권11호
• /
• pp.1084-1092
• /
• 2006

Shape optimization of airfoil in WIG craft has been performed by considering the ground effect. The WIG craft should satisfy various aerodynamic characteristics such as lift, lift to drag ratio, and static height stability. However, they show a strong trade-off phenomenon so that it is difficult to satisfy aerodynamic properties simultaneously. Optimization is carried out through the multi-objective genetic algorithm. A multi-objective optimization means that each objective is considered separately instead of weighting. Due to the trade-off, pareto sets and non-dominated solutions can be obtained instead of the unique solution. NACA0015 airfoil is considered as a baseline model, shapes of airfoil are parameterized and rebuilt with four-Bezier curves. There are eighteen design variables and three objective functions. The range of design variables and their resolutions are two primary keys for the successful optimization. By two preliminary optimizations, the variation can be reduced effectively. After thirty evolutions, the non-dominated pareto individuals of twenty seven are obtained. Pareto sets are all the set of possible and excellent solution across the design space. At any selections of the pareto set, these are no better solutions in all design space.

• Journal of Power Electronics
• /
• 제11권5호
• /
• pp.621-631
• /
• 2011

In space qualified DC-DC converters, optimization of the following electrical characteristics is of greater interest in comparison with other specifications; power loss/efficiency, output voltage ripple and volume/weight. The main goal of this paper is to present an appropriate solution for optimizing the above mentioned characteristics. For this purpose, a comprehensive power loss model of a DC-DC converter is fully developed. Proper models are also demonstrated for assessment of the output voltage ripple and the utilized transformer volume as the bulkiest component in a DC-DC converter. In order to provide a test bed for evaluation of the proposed models, a 50W push-pull DC-DC converter is designed and implemented. Finally, a novel cost function with three assigned weight functions is proposed in order to have a trade-off among the power loss, the output voltage ripple and the utilized transformer volume of the converter. The cost function is optimized for applications in which volume has the highest priority in comparison with power loss and ripple. The optimization results show that the transformer volume can be decreased by up to 51% and this result is verified by experimental results. The developed models and algorithms in this paper can be used for other DC-DC converter topologies with some minor modifications.

• Journal of Positioning, Navigation, and Timing
• /
• 제9권4호
• /
• pp.293-304
• /
• 2020

In this paper, various modulation techniques, including the legacy Global Navigation Satellite System (GNSS) signal modulation techniques, are introduced and the spectral characteristics and correlation characteristics of signals with various modulation techniques are analyzed based on numerical simulation. With the development of various GNSS services, the limited frequency band has become increasingly saturated, and issues of interoperability and compatibility have emerged in the new GNSS design. Since the efficient allocation of frequency resources is closely related to spectrum design, modulation techniques are one of the important signal design parameters of new signal design. Signal modulation techniques are closely related to various figure of merits (FoMs) as well as spectrum characteristic, and in some cases there is a complicated trade-off between FoMs. Thus, the FoMs associated with modulation technology should be analyzed and the best signal candidates should be chosen carefully via the trade-off analysis for FoMs. In this paper, we define the modulation technique based on Phase Shift Keying (PSK), Binary Offset Carrier (BOC) and Continuous Phase Modulation (CPM) for the design of KPS signals, and the FoMs of signals in terms of spectrum and correlation function are evaluated. Signals with various modulation techniques are implemented through a numerical simulation, and the relevant FoMs are analyzed.

• 대한기계학회논문집A
• /
• 제39권12호
• /
• pp.1199-1206
• /
• 2015

본 연구는 경유와 바이오디젤(대두유) 혼합연료의 디젤엔진 배기특성을 조사하였고, 연료 혼합비는 BD(biodiesel)3, BD5, BD20, BD50 및 BD100이며, 분사압력 조건을 400 bar, 600 bar, 800 bar, 1000 bar 및 1200 bar로 변화시켰다. 그리고 연료 혼합비 및 분사압력에 따른 엔진배출물인 NOx와 Soot의 정량적인 분석을 위해 통계학에 기초한 피어슨 상관계수와 스피어만 상관계수를 이하였다. 본 연구의 결과로서 실험변수인 혼합비와 분사압력에 대한 NOx 및 Soot 발생량의 피어슨 상관계수는 -0.811이며, 스피어만 상관계수는 -0.884로 NOx와 Soot 발생량 관계가 선형적이며, 이것은 trade-off관계를 나타낸다. 또한 각각의 분사압력 조건에서 피어슨 상관계수가 음의 상관 관계를 나타내며 이것은 NOx와 Soot 배출관계가 반비례적인 것을 나타낸다.

• 한국항공우주학회지
• /
• 제31권6호
• /
• pp.102-109
• /
• 2003

전투기에 장착되는 엔진의 수는 필요한 추력과 가용엔진에 의해 대부분 결정되어 왔으나, 엔진 기술의 발달로 임무성능-체계분석-경제성 등이 중요한 고려요소가 되었다. 베트남전과 걸프전의 실전사례 분석 결과 안전성과 취약성은 쌍발기가, 피격율은 단발기가 다소 우수한 것으로 평가되었다. F404-GE-400과 F-125 엔진을 장착한 초음속 경공격기를 설계하여, 엔진 수에 따른 비교연구를 수행하였다. 쌍발기 형상이 최대이륙중량 8%, Flyaway Cost 26%, LCC 13% 정도 컸으며, 단발기 형상이 기동-저속성능과 RM&S 성능이 다소 우수하였으나 큰 차이는 없었다. 전투기의 획득시 저급(Low) 전투기는 단발엔진을, 중급이상(Med.+)의 전투기는 쌍발엔진을 적용하는 것이 체계분석-경제성-운용개념 등을 고려한 여러 측면에서 유리할 것으로 판단된다.

• 인터넷정보학회논문지
• /
• 제25권2호
• /
• pp.1-9
• /
• 2024

연합학습은 보안 및 프라이버시 측면에서 중앙 집중식 방법보다 안전하도록 설계되었음에도 불구하고 여전히 많은 취약점을 내재한다. 적대적 공격(adversarial attack)을 수행하는 공격자는 신중하게 제작된 입력 데이터, 즉 적대적 예제(adversarial examples)를 클라이언트의 학습 데이터에 주입하여 딥러닝 모델을 의도적으로 조작하여 오분류를 유도한다. 이에 대한 보편적인 방어 전략은 이른바 적대적 학습(adversarial training)으로 적대적 예제들의 특성을 선제적으로 모델에 학습시키는 것이다. 기존의 연구에서는 모든 클라이언트가 적대적 공격 하에 있는 상황을 가정하는데 연합학습의 클라이언트 수가 매우 많음을 고려하면 실제와는 거리가 있다. 본 논문에서는 클라이언트의 일부가 공격 하에 있는 시나리오에서 적대적 학습의 양상을 실험적으로 살핀다. 우리는 실험을 통해 적대적 예제에 대한 분류 정확도가 증가하면 정상 샘플에 대한 분류 정확도의 감소하는 트레이드오프 관계를 가짐을 밝혔다. 이러한 트레이드오프 관계를 효과적으로 활용하기 위해 클라이언트가 자신이 공격받는지 여부에 따라 손실함수를 적응적으로 선택하여 적대적 학습을 수행하는 방법을 제시한다.

• 한국철도학회:학술대회논문집
• /
• 한국철도학회 1998년도 추계학술대회 논문집
• /
• pp.413-420
• /
• 1998

In this study, optimum design methodology for rail vehicle suspension characteristics is suggested. Three parameters, primary lateral/longitunal stiffness and secondary lateral stiffness, are selected as design parameters. critical speed, suspension stroke trade-off and derailment coefficient are selectee as performance constraints. The optimum parameters to maximize ride quality are evaluated under the constraints. Steady-state curiving model to be able to evaluate derailment coefficient is developed. The combined design procedure is developed to evaluate Three parameters at the same time.

• 전자공학회논문지A
• /
• 제32A권12호
• /
• pp.122-129
• /
• 1995

The most important feature of a-Si TFT is dense localized states such as dangling bonds which exist in tis bandgap. Electrons trapped by localized states dominate the potential distribution in the active a-Si region ,and influence the performance of a-Si TFT. In this paper, we describe the electrical characteristics of a-Si TFT with respect to trap distribution within bandgap, electron mobility and interface states using 2-Dimensional device simulator and compare the result of simulation with measurements. Using the mixed-mode simulator, we can predict the potential variation of pixel which causes residual image problem during the turn-off of a-Si TFT driving circuit. Therefore it is possible to consider trade-off between potential variation of pixel and turn-on current of a-Si TFT for the optimized driving circuit.

• The Journal of Asian Finance, Economics and Business
• /
• 제6권3호
• /
• pp.41-53
• /
• 2019

Building a target capital structure is one of the most important decisions in corporate financial management. The purpose of this article is to identify the determinants of capital structure and adjustment mechanism toward the target leverage. The partial adjustment model was applied on a sample of 306 non-financial companies listed on Vietnam stock exchange market during the period of 2008-2017. By the fixed effect model estimation method, the research results have discovered the factors of growth opportunities, firm size, tangible fixed assets and firm's unique characteristics have a positive effect on the target capital structure of enterprises. Besides, profitability and dividend payment have a negative effect on the target capital structure of enterprises. Accordingly, the research results show that the average adjustment speed toward target leverage of the firms is 90.03%. Research results also demonstrate firms have higher or lower debt ratio than the target debt ratio, capital surplus or capital deficit also have an impact on the adjustment rate toward the target capital structure. The research results are consistent with the Dynamic Trade-off Theory. From this result, this article has provided policy implications for non-financial companies listed on Vietnam's stock market in building a reasonable target capital structure according to operating timeline to maximize enterprise value.

• 정보처리학회논문지C
• /
• 제11C권6호
• /
• pp.863-870
• /
• 2004

최근들어 UMTS(Universal Mobile Telecommunication System)는 3세대 무선망의 표준으로 인식되고 있고 DSCH(Down-link Shared Chanel)는 UMTS에서 매우 효율적인 다중화 방안으로 고려되고 있다. 그러므로 UMTS에서의 DSCH를 위한 효과적인 패킷스케줄링 알고리즘 개발에 대한 연구가 필수적이다. 본 논문은 UMTS의 DSCH를 위한 개선된 패킷스케줄링 알고리즘에 대하여 기술한다. 이 알고리즘은 최대 링크 이용율을 위하여 이동호스트채널상태를 고려하며, 동시에 장기적으로 공정성을 얻기 위해 3등급 트래픽에서의 지연 한계치와 4등급 트래픽의 처리율을 함께 이용한다. 알고리즘의 타당성을 보이기 위하여 PARSEC을 이용한 모의실험을 수행하였다. 모의실험 결과는 이 알고리즘이 무선링크 이용율과 공정성 두 가지 측면에서 적적한 타협(trade-off)을 통하여 개선된 결과를 얻을 수 있음을 보여준다.