Earticle

일반적으로 게임 디자인 시에 웹과 같은 매체를 통해 구하거나 디자이너들이 디자인 툴로 제작한 3차원 기하 모델은 polygon soup 형태가 대부분이다. 따라서 이러한 polygon soup 모델은 일반적으로 완전한 매시로 가정해서 적용하는 여러 기법들을 적용할 수 없다. 이러한 문제를 해결하고 보다 매끈한 곡면을 얻기 위해서 MLS(Moving Least Squares) 방법을 점 단위가 아니라 삼각형 면 단위로 확장한 적분형 MLS 기법이 제안되었다. 그러나 이 기법은 본질적으로 전역(global) 계산의 한계로 인한 계산 속도의 한계가 불가피한 특징이 있었다. 본 논문에서는 분석적 해를 바탕으로 한 전역 계산 속도를 보다 가속화하기 위한 GPU 기반 병렬 기법을 제안하고 또한 기존 논문에서 충분히 논의하지 않았던 적분형 MLS 기법과 일반 MLS 기법의 차이점을 설명한다. 특히 GPU를 통한 가속 결과, 연산 정밀도의 감소없이 CPU 코어 1개로 계산하는 경우보다 평균 250배의 속도 향상을 얻을 수 있었다.

Most 3D models which are easily available on the Internet or designed by designers are polygon soups - they do not have topology information; some triangles are missing; or there exist some T-junctions. Due to this issue, it is often impossible to apply advanced geometry techniques to those polygon soups because most advanced methods assume that the 3D models are perfect manifold meshes. To address this problem, an integral MLS method which calculates integration, not summation over sample points to generate smooth implicit surfaces. However, those methods basically require global calculations for all the triangles so that it is very hard to accelerate the operations only with one CPU core. In this paper, we present a GPU-based parallelization technique for the analytic approach of the integral MLS method. Also, we discuss the difference between integral MLS and conventional one which was not fully explained in other papers. In our experiment, using GPU, we achieve that 250-fold speedup over the single CPU core.

Sphere decoding is a very powerful algorithm in searching the optimal solution of multiple input and multiple output systems. However, it cannot perform in parallel directly. Sphere decoding can be depicted as searching in a tree. In this paper, we propose a parallel mode of the sphere decoding algorithm. We proposed that the searching tree can be partitioned into several sub-trees. The searching is divided into two stages. In the first stage, the partial Euclidean distances of sub-tree root nodes are calculated. In the second stage, several sub-trees perform their searching simultaneously. The Euclidean distance of the early finished sub-tree helps to reduce the calculation in the later finished sub-trees search. Simulation results sho

Genome-wide association study (GWAS) may be characterized by relatively simple calculation for individual single nucleotide polymorphism (SNP) or SNP combinations and rapidly increasing number of such calculation when considering higher order interactions between SNPs. In principle, simultaneous evaluation of association measure for multiple SNP or SNP combinations should expedite the whole study, which coincides with the definition of the parallel computing. Distributing the job to connected and controllable many cpu’s has been called as the clustering. However as the cpu integration technology develops, a single physical cpu can have multiple cores, over which a job may be distributed and executed concurrently. More recently, general purpose graphics processing unit (GPGPU) is drawing attention, which may be characterized by much more cores with less computing powers. Serial codes should be converted into parallel version so as to take advantage of the multi-core central processing unit (CPU) and GPGPU. Using specialized directives or pragmas called openMP and openACC, performance of parallel computing is examined on genomic data in various computing conditions.

This paper introduces a novel parallel shift operator finite-difference time-domain (SO-FDTD) method for plasma in the dispersive media. We calculate the interaction between the electromagnetic wave of various frequencies and non-magnetized plasma by using the parallel SO-FDTD method. Then, we compare the results，which are calculated with serial and parallel SO-FDTD executions to obtain the speedup ratio and validate the parallel execution. We conclude that the parallel SO method has almost the same precision as the serial SO method, while the parallel approach expands the scope of memory and reduces the CPU time.

An efficient inverse design technique based on the MGM (Modified Garabedian-McFadden) method has been developed. The 2-D Navier-Stokes equations are solved for obtaining the surface pressure distributions and coupled with the MGM method to perform the inverse design. The solver is parallelized by using the domain decomposition method and the standard MPI library for communications between the processors. The MGM method is a residual-correction technique, in which the residuals are the difference between the desired and the computed pressure distribution. The developed code was applied to several airfoil shapes and the axial blade. It has been found that they are well converged to their target pressure distribution.

In this paper, the equations of motions for planar multibody dynamics are established for considering the parallel programming based on GPU. Cartesian coordinates are used to formulate the equations of motion and implicit integration method called HHT-alpha is employed. Open chain multibody system is considered for computer simulation. CUDA toolkit is employed for establishing the GPU parallel programming. The exactness of the analysis is verified from the comparison with ADAMS. The results from parallel computing based on GPU are compared with the results from the sequential programming based on CPU in terms of calculation time. The multiple pendulum with bodies and joints is employed for the computer simulation. In the pendulum system that has 290 bodies, the parallel program indicates an improved efficiency of about 25.5 second(15.5% improvement). It is noted that the larger the size of system is, the time efficiency is better.

주파수 영역 유한요소 파동방정식의 3차원 모델링은 거대한 크기의 산재행렬(sparse matrix)인 임피던스 행렬을 풀어야 한다. 이러한 이유 때문에 파동방정식의 3차원 모델링은 주로 시간 영역에서 이루어지고 있다. 이 연구는 주파수 영역 파동방정식의 유한요소 3차원 모델링 연구의 일환으로 라플라스 영역에서 1개 주파수에 대한 파동방정식 해를 이용하여 주시와 진폭을 계산할 수 있는 SWEET(Suppressed Wave Equation Estimation of Traveltime) 알고리즘과 병렬 유한요소 솔버를 결합하여 주파수 영역 3차원 모델링을 시도 하였다. 이렇게 계산된 주시와 진폭은 파선이론에 기반하여 계산된 주시와 진폭과 달리 급경사 구조 또는 수평 속도의 비가 큰 곳에서도 정확하게 계산되며, Kirchhoff 구조보정에 유용하게 사용될 수 있다. 연구의 결과를 검증하기 위하여 SEG/EAGE 3D 암염 모델의 주시와 진폭 계산에 적용하여 이를 검증하였다.

In order to calculate 3-dimensional wavefield using finite-element method in frequency domain, we must factor so huge sparse impedance matrix. Because of difficulties of handling of this huge impedance matrix, 3-dimensional wave equation modeling is conducted mainly in time domain. In this study, we simulate the 3-D wavefield using finite-element method in Laplace domain by combining high-performance parallel finite-element solver and SWEET (Suppressed Wave Equation Estimation of Traveltime) algorithm which can calculate the traveltime and the amplitude. To verify this combination, we applied it to the SEG/EAGE 3D salt model in serial and parallel computing environments.

항내 혹은 협수로와 같은 제한수역에서 근접하여 항해중인 두 선박간의 상호 유체력 및 모멘트가 선박조종운동에 상당히 크게 영향을 준다는 것은 잘 알려져 있다. 두 선박간의 상호간섭력은 종방향 거리, 횡방향 거리 및 두 선박의 속도의 함수로서 가정 될 수 있다. 본 연구에서는 근접 항해중인 추월선박과 피 추월선박간의 상호유체력을 계산하고, 두 선박간의 횡방향 및 종방향 거리와 속도, 그리고 타사용으로 인한 접근상황 및 충돌시간에 대해서 검토, 고찰하였다.

It is well known tint the ship manoeuvring motion is greatly affected by hydrodynamic forces and moments acting between two vessels passing too close to each other in confined waters, such as in a harbour or narrow channel. This interaction between two vessels could be assumed to be the functions of the longitudinal distance, transverse distance and their speeds. The aim of this study is to calculate the interaction between two vessels passing close to each other on parallel courses by simulation, and to estimate the effect of rudder action and time of collision through simulation under the condition of various longitudinal distances and different speed-ratios of the two vessels.

원격탐사는 넓은 지역을 직접 접촉하지 않고 정보를 취득할 수 있고 다양한 분야에 적용할 수 있음으로써 급속히 발전하게 되었다. 이에 따라 위성의 제원 또한 원격탐사의 발전과 함께 급속한 발전을 이루게 되었다. 이러한 이유로 여러 분야에서 활용에 관한 연구가 활발히 이루어지고 있다. 현재 활용에 관한 연구는 활발히 이루어지고 있지만, 자료처리에 관련된 연구가 부족한 실정이다. 예전보다 인공위성의 제원이 발전하면서 많은 양의 정보 획득이 가능해진 것과 동시에 데이터 크기 또한 매우 커졌다. 이는 과거에 비해 자료의 처리속도가 저하된다는 단점이 존재한다. 따라서 본 논문에서는 병렬 처리의 한 가지 기법인 NVIDIA에서 제공하고 있는 CUDA (Compute Unified Device Architecture) 라이브러리를 활용하여 위성영상 자료처리 성능의 최적화를 목적으로 하고 있다. 본 연구의 순서는 다음과 같다. 다목적실용위성(Korea Multi-Purpose Satellite, KOMPSAT)의 영상을 크기를 기준으로 5가지 Type으로 나눈다. 이렇게 나누어진 영상을 원격탐사 분야의 한 가지 방법인 NDVI (Normalized Difference Vegetation Index)로 구현한다. 이때 CPU (Central Processing Unit, 중앙처리장치) 기반과 GPU (Graphic Processing Unit, 그래픽처리장치) 기반의 두 가지 방법과 상용 소프트웨어인 ArcMap을 이용하여 NDVI를 구현한다. 그리고 동일한 영상 유무를 판단하기 위해 구현된 결과 영상들을 히스토그램과 시각적으로 비교하고 CPU 버전과 GPU 버전의 처리속도를 비교 분석하였다. 연구결과 CPU 버전과 GPU 버전의 결과 영상은 ArcMap으로 구현한 영상과 시각적 그리고 히스토그램 비교를 통해 같은 결과를 나타내어 NDVI 코드는 올바르게 구현되었으며, 처리속도는 CPU보다 GPU가 약 5배 정도 빠른 것으로 확인하였다. 본 연구에서 병렬 처리의 한 기법인 CUDA 라이브러리를 활용하여 위성영상 자료처리 성능을 향상시킬 수 있었으며, 향후 NDVI와 같은 단순한 픽셀 연산 이외에도 다양한 원격탐사 기법의 적용이 필요할 것으로 사료된다.

Remote sensing allows acquisition of information across a large area without contacting objects, and has thus been rapidly developed by application to different areas. Thus, with the development of remote sensing, satellites are able to rapidly advance in terms of their image resolution. As a result, satellites that use remote sensing have been applied to conduct research across many areas of the world. However, while research on remote sensing is being implemented across various areas, research on data processing is presently insufficient; that is, as satellite resources are further developed, data processing continues to lag behind. Accordingly, this paper discusses plans to maximize the performance of satellite image processing by utilizing the CUDA(Compute Unified Device Architecture) Library of NVIDIA, a parallel processing technique. The discussion in this paper proceeds as follows. First, standard KOMPSAT(Korea Multi-Purpose Satellite) images of various sizes are subdivided into five types. NDVI(Normalized Difference Vegetation Index) is implemented to the subdivided images. Next, ArcMap and the two techniques, each based on CPU or GPU, are used to implement NDVI. The histograms of each image are then compared after each implementation to analyze the different processing speeds when using CPU and GPU. The results indicate that both the CPU version and GPU version images are equal with the ArcMap images, and after the histogram comparison, the NDVI code was correctly implemented. In terms of the processing speed, GPU showed 5 times faster results than CPU. Accordingly, this research shows that a parallel processing technique using CUDA Library can enhance the data processing speed of satellites images, and that this data processing benefits from multiple advanced remote sensing techniques as compared to a simple pixel computation like NDVI.

This paper described the unbalanced voltage on the tertiary bus of a single Phase auto transformer in the case of parallel operation with different manufacturer at each Phase. The unbalanced capacitances between primary to secondary winding, secondary to tertiary winding and primary to tertiary winding makes unbalanced bus voltage in the tertiary bus side. The unbalanced voltage let the surge arrester to operate in the power frequency range, and it causes the arrester to burn out. The failure of the arrester at one phase makes line to ground fault, which lead to the surge arrester failure of the other two phase on the tertiary bus.

본 연구는 다중채널코일과 병렬영상기법 이용 시 관행적으로 사용하고 있는 두영역측정법을 사용한 신호대잡음비 측정의 문제점을 알아보고자 하였다. 두영역측정법 사용 시 3가지 전제 조건에 만족하는 단일채널구상코일을 이용한 기준 SNR을 산출한 후, 다중채널코일과 병렬영상기법 이용 시 공신력 있는 기관에서 권고하는 방법과 전제 조건을 고려하지 않고 사용되어 문제가 되는 두영역측정법을 사용하여 SNR을 산출한 후 비교하였다. 다중채널코일과 병렬영상기법 이용 한 두영역측정법이 가장 높은 상대표준편차를 보여 낮은 정밀도를 나타내었고, ROI 위치에 따른 SNR 차이도 매우 높아 공간적으로 잡음의 분포가 균일하지 않음을 알 수 있었다. Blend-Altman plot 통한 95% 신뢰구간 간격도 가장 넓어 기준이 되는 단일채널구상코일을 이용한 두영역측정법 대비 일치도가 낮음을 알 수 있었다. 본 연구는 동일한 영상획득조건에서 자기공명영상장치 성능평가 시험의 표준이 되는 AAPM 방법과, 신호 영역 내에서 잡음 레벨을 정확히 결정할 수 있는 NEMA 방법, 그리고 자기공명영상장치의 제조사가 권고하는 방법을 직접적으로 비교함으로써, 연구자들이 간과할 수 있는 전제조건을 만족하지 않는 다중채널코일과 병렬영상기법 이용 시, 두영역측정법을 사용한 신호대잡음비의 부정확한 문제점을 정량적으로 증명하였다는데 큰 의의가 있다.

The purpose of this study was to investigate the problems of a signal to noise ratio measurement using a two region measurement method that is conventionally used when using a multi-channel coil and a parallel imaging technique. As a research method, after calculating the standard SNR using a single channel head coil of which coil satisfies three preconditions when using a two region measurement method, we made comparisons and evaluations after calculating an SNR by using a two region measurement method of which method is problematic because it is used without considering the methods recommended by reputable organizations and the preconditions at the time of using a multi-channel coil and a parallel imaging technique. We found that a two region measurement method using a multi-channel coil and a parallel imaging technique shows the highest relative standard deviation, and thus shows a low degree of precision. In addition, we found out that the difference of SNR according to ROI location was very high, and thus a spatial noise distribution was not uniform. Also, 95% confidence interval through Blend-Altman plot is the widest, and thus the conformity degree with a two region measurement method using the standard single channel head coil is low. By directly comparing an AAPM method, which serves as a standard of a performance evaluation test of a magnetic resonance imaging device under the same image acquisition conditions, an NEMA method which can accurately determine the noise level in a signal region and the methods recommended by manufacturers of a magnetic resonance imaging device, there is a significance in that we quantitatively verified the inaccurate problems of a signal to noise ratio using a two region measurement method when using a multi-channel coil and a parallel imaging technique of which method does not satisfy the preconditions that researchers could overlook.