• 한국시뮬레이션학회논문지
• /
• 제18권2호
• /
• pp.57-64
• /
• 2009

물리적 기반의 다이내믹 시뮬레이션에 있어서 안정적이고 효율적인 제한 시스템은 매우 중요한 요소 중 하나이다. 본 논문은 기존에 널리 사용되고 있는 제한 시스템들(Lagrange Multiplier method, Baumgarte stabilization method, Post-stabilization Method, Implicit constraint enforcement method, Fast projection method)에 대한 분석과 평가를 통해 제한 시스템을 사용하고자 하는 사용자들에게 적절한 선택을 할 수 있는 지침을 제공하고자 한다. 본 논문은 기존의 제한 방법들에 대한 수학적 수식들이 설명되어 있고, 제한 오차 비교, 계산 비용, 동적 움직임 분석 등을 통해 기존 제한 시스템들 각각에 대한 평가를 제공한다.

• 한국재료학회지
• /
• 제20권9호
• /
• pp.478-482
• /
• 2010

A series of molecular dynamic (MD), finite element (FE) and ab initio simulations are carried out to establish suitable modeling schemes for the continuum-based analysis of aluminum matrix nanocomposites reinforced with carbon nanotubes (CNTs). From a comparison of the MD with FE models and inferences based on bond structures and electron distributions, we propose that the effective thickness of a CNT wall for its continuum representation should be related to the graphitic inter-planar spacing of 3.4${\AA}$. We also show that shell element representation of a CNT structure in the FE models properly simulated the carbon-carbon covalent bonding and long-range interactions in terms of the load-displacement behaviors. Estimation of the effective interfacial elastic properties by ab initio simulations showed that the in-plane interfacial bond strength is negligibly weaker than the normal counterpart due to the nature of the weak secondary bonding at the CNT-Al interface. Therefore, we suggest that a third-phase solid element representation of the CNT-Al interface in nanocomposites is not physically meaningful and that spring or bar element representation of the weak interfacial bonding would be more appropriate as in the cases of polymer matrix counterparts. The possibility of treating the interface as a simply contacted phase boundary is also discussed.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제2권1호
• /
• pp.51-62
• /
• 2008

This paper proposes a second-order implicit constraint enforcement method which yields enhanced controllability compared to a first-order implicit constraints enforcement method. Although the proposed method requires solving a linear system twice, it yields superior accuracy from the constraints error perspective and guarantees the precise and natural movement of objects, in contrast to the first-order method. Thus, the proposed method is the most suitable for exact prediction simulations. This paper describes the numerical formulation of second-order implicit constraints enforcement. To prove its superiority, the proposed method is compared with the firstorder method using a simple two-link simulation. In this paper, there is a reasonable discussion about the comparison of constraints error and the analysis of dynamic behavior using kinetic energy and potential energy.

• 한국컴퓨터그래픽스학회논문지
• /
• 제30권2호
• /
• pp.1-9
• /
• 2024

본 논문에서는 밀도 데이터로부터 다양한 벡터장 패턴을 시각화하는 새로운 방법을 제안한다. 이를 위해 물리 기반 시뮬레이션과 기하학적 처리에서 사용되는 이동최소제곱(Moving least squares, MLS)을 이용한다. 하지만 일반적인 MLS는 벡터기반의 제약조건을 통해 고차 보간되기 때문에 밀도의 특성을 고려하지 못한다. 본 논문에서는 입력 데이터에 내포되어 있는 밀도의 특성을 효율적으로 고려하기 위해 몬테카를로 기반의 가중치를 MLS에 통합하여 다양한 형태의 백터장을 표현할 수 있도록 알고리즘을 설계한다. 결과적으로 일반적인 MLS와 발산제약 기반의 MLS 같은 기존 기법으로는 표현하기 힘든 디테일한 벡터장을 실험을 통해 보여준다.

• 한국인공지능학회지
• /
• 제10권2호
• /
• pp.7-11
• /
• 2022

OLED displays cannot be used permanently due to burn-in or generation of dark spots due to degradation. Therefore, the time when the display can operate normally is very important. It is close to impossible to physically measure the time when the display operates normally. Therefore, the time that works normally should be predicted in a way other than a physical way. Therefore, if you do computer simulations based on artificial intelligence, you can increase the accuracy of prediction by saving time and continuous learning. Therefore, if we do computer simulations based on artificial intelligence, we can increase the accuracy of prediction by saving time and continuous learning. In this paper, a dataset in the form of development from generation to diffusion of dark spots, which is one of the causes related to the life of OLED, was generated by applying the finite element method. The dark spots were generated in nine conditions, such as 0.1 to 2.0 ㎛ with the size of pinholes, the number was 10 to 100, and 50% with water content. The learning data created in this way may be a criterion for generating an artificial intelligence-based dataset.

• Journal of Information Processing Systems
• /
• 제8권4호
• /
• pp.713-720
• /
• 2012

Unlike FEM (Finite Element Method), which provides an accurate deformation of soft objects, FFD (Free Form Deformation) based methods have been widely used for a quick and responsive representation of deformable objects in real-time applications such as computer games, animations, or simulations. The FFD-AABB (Free Form Deformation Axis Aligned Bounding Box) algorithm was also suggested to address the collision handling problems between deformable objects at an interactive rate. This paper proposes an enhanced FFD-AABB algorithm to improve the frame rate of simulation by adding the bounding sphere based collision test between 3D deformable objects. We provide a comparative analysis with previous methods and the result of proposed method shows about an 85% performance improvement.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1991년도 하계학술대회 논문집
• /
• pp.767-770
• /
• 1991

For the control of redundant manipulators, conventional dynamic control methods of local torque optimization showed the instability which resulted in physically unrealizable torque requirements. In this paper, a new dynamic control method which is based on the concept of aspects is proposed. The proposed method starts with the basic understanding of the minors in the Jacobian matrix. It was shown by computer simulations that the proposed method demonstrates a drastic reduction of torque loadings at the joints in the tracking motion of a long trajectory, and thus guarantees the stability of joint torque.

• Journal of Electrical Engineering and Technology
• /
• 제9권6호
• /
• pp.2079-2088
• /
• 2014

In this paper, we propose new physically based threshold voltage models for short channel Surrounding Gate Silicon Nanowire Transistor with two different geometries. The model explores the impact of various device parameters like silicon film thickness, film height, film width, gate oxide thickness, and drain bias on the threshold voltage behavior of a cylindrical surrounding gate and rectangular surrounding gate nanowire MOSFET. Threshold voltage roll-off and DIBL characteristics of these devices are also studied. Proposed models are clearly validated by comparing the simulations with the TCAD simulation for a wide range of device geometries.

• 한국멀티미디어학회논문지
• /
• 제9권11호
• /
• pp.1515-1528
• /
• 2006

물리기반 유체 애니메이션 시스템들이 시각 특수효과 산업계에서 빠르게 발전하고 있고, 이를 이용해 매우 높은 화질의 영상을 제작하는 것이 가능하게 되었다. 그러나 컴퓨터 게임과 같은 실시간 응용 분야의 경우, 화질보다는 시뮬레이션 속도가 더 중요한 문제이다. 본 논문은 프로그래머블 그래픽스 파이프라인을 이용하여 유체에 대한 애니메이션을 수행하는 실시간 기법에 대해 설명한다. 두 개의 키-프레임이 주어졌을 때, 본 기법을 이용하여 원시 프레임으로부터 목적 프레임으로 변하는 연속 영상을 대화식으로 생성할 수 있음을 보인다.

• International Journal of Contents
• /
• 제14권4호
• /
• pp.24-29
• /
• 2018

EEG-based brain-computer interfaces has focused on explicitly expressed intentions to assist physically impaired patients. For EEG-based-computer interfaces to function effectively, it should be able to understand users' implicit information. Since it is hard to gather EEG signals of human brains, we do not have enough training data which are essential for proper classification performance of implicit intention. In this paper, we improve the subject independent classification of implicit intention through the generation of additional training data. In the first stage, we perform the PCA (principal component analysis) of training data in a bid to remove redundant components in the components within the input data. After the dimension reduction by PCA, we train ICA (independent component analysis) network whose outputs are statistically independent. We can get additional training data by adding Gaussian noises to ICA outputs and projecting them to input data domain. Through simulations with EEG data provided by CNSL, KAIST, we improve the classification performance from 65.05% to 66.69% with Gamma components. The proposed sample generation method can be applied to any machine learning problem with fewer samples.