• 제어로봇시스템학회논문지
• /
• 제15권4호
• /
• pp.451-455
• /
• 2009

This paper provides a comparison of global path planning method in single string by using pulled and pushed SOFM (Self-Organizing Feature Map) which is a method among a number of neural network. The self-organizing feature map uses a randomized small valued initial-weight-vectors, selects the neuron whose weight vector best matches input as the winning neuron, and trains the weight vectors such that neurons within the activity bubble are move toward the input vector. On the other hand, the modified SOFM method in this research uses a predetermined initial weight vectors of the one dimensional string, gives the systematic input vector whose position best matches obstacles, and trains the weight vectors such that neurons within the activity bubble are move toward or reverse the input vector, by rising a pulled- or a pushed-SOFM. According to simulation results one can conclude that the modified neural networks in single string are useful tool for the global path planning problem of a mobile robot. In comparison of the number of iteration for converging to the solution the pushed-SOFM is more useful than the pulled-SOFM in global path planning for mobile robot.

• 대한기계학회논문집
• /
• 제18권4호
• /
• pp.958-966
• /
• 1994

The postbuckling compressive strengths of $[0/90/\pm\theta]_s$ composite laminated cylindrical panels with various fiber angles and width-to-length ratios are characterized by the nonlinear finite element method. For the iteration and load-increment along the postbuckling equilibrium path a modified arc-length method in which the effect of failure can be considered is introduced. In the progressive failure analysis the maximum stress criterion and complete unloading model are used. Present finite element results show good agreement with experiments for $[0_3/90]_s$ cylindrical panel and $[0/\pm45/90/]_s$ plate. The postbuckling compressive strength of $[0/90/\pm\theta]_s$ composite laminated cylindrical panel is independent of the initial buckling stress but high in the panel with large value of the bending stiffness in axial direction. In the several cylindrical panels, it is observed that the prebuckling compressive failures occur and result into the collapse before the buckling.

• 한국컴퓨터정보학회논문지
• /
• 제10권2호
• /
• pp.1-10
• /
• 2005

본 논문에서는 연속적인 레인지(range) 영상 자료로부터 동작 벡터를 추출하는 새로운 블록 정합 알고리즘을 제안한다. 본 논문에서 제안된 알고리즘은 단일 특징을 사용하지 않고 다중 특징인 명암값, 색상, 레인지 특징의 세 가지 특징을 통합한 정합 유사 함수를 정의하며, 엔트로피를 이용하여 각 특징의 기여도를 구한 후 이를 가중치의 형태로 정합 유사 함수에 적용한다. 그리고 제안된 알고리즘은 고정된 블록 템플릿을 사용하지 않고 가변적인 크기의 블록 템플릿을 사용한다. 제안한 블록 정합에서는 먼저 작은 정합 템플릿으로 블록 정합을 시작한다. 만일 정합 정도가 좋지 않으면 정합 템플릿의 크기를 조금 확장한 후 본 논문에서 정의한 정합기준이 만족하거나 사전에 정의된 최대 블록 크기에 도달할 때까지 블록정합을 반복한다. 실험에서는 본 논문에서 제안한 블록 정합 알고리즘과 기존의 다른 알고리즘의 성능을 비교 분석하여 제안한 알고리즘의 우수함을 보인다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제9권6호
• /
• pp.2302-2316
• /
• 2015

We propose an efficient method of extracting targets within a region of interest in non-homogeneous infrared images by using a principal component analysis (PCA) plane and adaptive Gaussian kernel. Existing approaches for extracting targets have been limited to using only the intensity values of the pixels in a target region. However, it is difficult to extract the target regions effectively because the intensity values of the target region are mixed with the background intensity values. To overcome this problem, we propose a novel PCA based approach consisting of three steps. In the first step, we apply a PCA technique minimizing the total least-square errors of an IR image. In the second step, we generate a binary image that consists of pixels with higher values than the plane, and then calculate the second derivative of the sum of the square errors (SDSSE). In the final step, an iteration is performed until the convergence criteria is met, including the SDSSE, angle and labeling value. Therefore, a Gaussian kernel is weighted in addition to the PCA plane with the non-removed data from the previous step. Experimental results show that the proposed method achieves better segmentation performance than the existing method.

• 전산구조공학
• /
• 제9권3호
• /
• pp.209-216
• /
• 1996

본 논문의 목적은 단층래티스 돔의 안정해석 기법을 제안하고, 다양한 조건하에서 단층 래티스 돔이 갖는 좌굴특성을 규명하는데 있다. 기하학적 비선형 평형경로의 탐색 및 좌굴점 그리고 분기경로의 방향 등을 수직적으로 계산하기 위하여 호장법(Arc-Length Method)을 이용하였으며, 부재의 좌굴가능성을 판단하기 위하여 에너지밀도함수를 제안하였다. 강절점을 갖는 구조물의 거동 특성을 규명하기 위하여 3종류의 비선형 강성행렬을 유도하여 해석하였으며, semi-rigid절점을 갖는 구조물을 해석하기 위하여 스프링모델을 제안하고, 부재의 세장비, 반개각, 경계조건 및 다양한 하중조건을 파라메터로하여 좌굴특성을 규명하였다.

• Bulletin of the Korean Chemical Society
• /
• 제24권3호
• /
• pp.279-284
• /
• 2003

The One Pass Method (OPM) previously presented for the identification of single input single output systems is used to estimate the parameters of a Decentralized Control System (DCS). The OPM is a linear and therefore a simple estimation method. All of the calculations are performed in one pass, and no initial parameter guess, iteration, or powerful search methods are required. These features are of interest especially when the parameters of multi input-output model are estimated. The benefits of the OPM are revealed by comparing its results against those of two recently published methods based on pulse testing. The comparison is performed using two databases from the literature. These databases include single and multi input-output process transfer functions and relevant disturbances. The closed loop responses of these processes are roughly captured by the previous methods, whereas the OPM gives much more accurate results. If the parameters of a DCS are estimated, the OPM yields the same results in multi or single structure implementation. This is a novel feature, which indicates that the OPM is a convenient and practice method for the parameter estimation of multivariable DCSs.

• Nuclear Engineering and Technology
• /
• 제52권11호
• /
• pp.2667-2677
• /
• 2020

A domain decomposition (DD) scheme for GPU-based Monte Carlo (MC) calculation which is essential for whole-core depletion is introduced within the framework of the modified history-based tracking algorithm. Since GPU-offloaded MC calculations suffer from limited memory capacity, employing DDMC is inevitable for the simulation of depleted cores which require large storage to save hundreds of newly generated isotopes. First, an automated domain decomposition algorithm named wheel clustering is devised such that each subdomain contains nearly the same number of fuel assemblies. Second, an innerouter iteration algorithm allowing overlapped computation and communication is introduced which enables boundary neutron transactions during the tracking of interior neutrons. Third, a bank update scheme which is to include the boundary sources in a way to be adequate to the peculiar data structures of the GPU-based neutron tracking algorithm is presented. The verification and demonstration of the DDMC method are done for 3D full-core problems: APR1400 fresh core and a mock-up depleted core. It is confirmed that the DDMC method performs comparably with the standard MC method, and that the domain decomposition scheme is essential to carry out full 3D MC depletion calculations with limited GPU memory capacities.

• Journal of Magnetics
• /
• 제16권2호
• /
• pp.181-185
• /
• 2011

To improve trapped field characteristics of a high temperature superconducting (HTS) bulk, a technique to implement artificial holes has been studied. The artificial holes, filled up with epoxy or metal, may provide better cooling channel and enhance mechanical strength of the HTS bulk. Although many useful researches based on experiments have been reported, a numerical approach is still limited because of several reasons that include: 1) highly non-linear electromagnetic properties of HTS; and 2) difficulty in modeling of randomly scattered "small" artificial holes. In this paper, a 2-D finite element method with iteration is adopted to analyze trapped field characteristics of HTS bulk with artificial holes. The validity of the calculation is verified by comparison between measurement and calculation of a trapped field in a $40{\times}40\;mm$ square and 3.1 mm thick HTS bulk having 16 artificial holes with diameter of 0.7 mm. The effects of sizes and array patterns of artificial holes on distribution of trapped field within HTS bulk are numerically investigated using suggested method.

• 대한전자공학회논문지SD
• /
• 제37권12호
• /
• pp.24-28
• /
• 2000

본 논문에서는 레벨셋 방법을 이용한 표면 전진기를 사용하여 반도체 공정에서 3차원 증착 프로파일을 계산하는 방법을 보고한다. 레벨셋 방법을 이용하여 효율적으로 표면의 전진, 후퇴를 모델링 하기 위하여 반복법에 의한 초기화 방법을 새로이 개발하였다. 또한, 각각의 위치에서의 증착 및 식각률을 계산하기 위하여 소스에서 표면까지의 가시도 계산 방법과 표면 반사 모델을 개발하였다. 3차원 표면 전진기로 그림자 효과와 표면 반사가 고려된 계산을 수행하였다. 흡착 계수가 1인 경우와 0.3인 경우를 비교하여 반사, 스퍼터링에 의한 증착 프로파일의 변화를 비교하였다. 그리고 종횡비에 따른 증착률의 차이를 그림자 효과에 의한 가시각의 차이로 비교하였다.

• Structural Engineering and Mechanics
• /
• 제8권3호
• /
• pp.233-242
• /
• 1999

In the present study, the elasto-plastic analysis of prismatic plate structures subjected to pure bending is carried out using the finite strip method. The end cross-sections of the structure are assumed to remain plane during deformation, and the compatibility along corner lines is ensured by choosing proper displacement functions. The effects of both the initial geometrical imperfections and residual stresses due to fabrication are included in the combined geometrically and materially nonlinear simulation. The von-Mises yield criterion and the Prandtl-Reuss flow theory of plasticity are applied in modelling the elasto-plastic behavior of material. Newton-Raphson iterations are carried out as the rotation of the end cross sections of the structure is increased step by step. The parameter representing the overall axial strain of structure is adjusted constantly during the iteration process in order to eliminate the resulting overall axial force on any cross-section of the structure in correspondence with the assumption of zero axial force in pure bending. Several numerical examples are presented to validate the present method and to investigate the effects of some material and geometrical parameters.