• Journal of Electrical Engineering and information Science
• /
• 제2권5호
• /
• pp.79-88
• /
• 1997

This paper introduces a new energy function, as maximum a posteriori(MAP) estimate of binocular disparity, that can deal with both random dot stereo-gram(RDS) and natural scenes. The energy function uses phase-magnitude as features to detect only the shift for a pair of corrupted conjugate images. Also we adopted Fleet singularity that effectively detects unstable areas of image plant and thus eliminates in advance error-prone stereo mathcing. The multi-scale concept is applied to the multi laser architecture that can search the solutions systematically from coarse to fine details and thereby avoids drastically the local minima. Using mean field approximation, we obtained a compact representation that is suitable for fast computation. In this manner, the energy function satisfies major natural constraints and requirements for implementing parallel relaxation. As an experiment, the proposed algorithm is applied to RDS and natural stereo images. As a result we will see that it reveals good performance in terms of recognition errors, parallel implementation, and noise characteristics.

• 정보처리학회논문지A
• /
• 제8A권4호
• /
• pp.455-460
• /
• 2001

어떤 응용프로그램에서는 계산구조가 프로그램의 수행도중 동적으로 변한다. 이런 경우 정적으로 태스크를 분할하고 할당하는 것은 병렬컴퓨터에서 높은 성능을 얻는데 충분하지 못하다. 이 논문에서는 동적으로 변하는 트리구조를 가진 계산을 프로세서들에 효율적으로 분배하는 부하분산 알고리즘을 소개한다. 이 알고리즘의 메쉬구조상에서의 구현기법이 소개되고 복잡도가 분석된다. 실험을 통하여 이 알고리즘이 좋은 성능을 나타내는 것을 보인다.

• Journal of information and communication convergence engineering
• /
• 제19권3호
• /
• pp.155-160
• /
• 2021

Differential evolution is an efficient algorithm for solving continuous optimization problems. However, its performance deteriorates rapidly, and the runtime increases exponentially when differential evolution is applied for solving large-scale optimization problems. Hence, a novel cooperative coevolution differential evolution based on Spark (known as SparkDECC) is proposed. The divide-and-conquer strategy is used in SparkDECC. First, the large-scale problem is decomposed into several low-dimensional subproblems using the random grouping strategy. Subsequently, each subproblem can be addressed in a parallel manner by exploiting the parallel computation capability of the resilient distributed datasets model in Spark. Finally, the optimal solution of the entire problem is obtained using the cooperation mechanism. The experimental results on 13 high-benchmark functions show that the new algorithm performs well in terms of speedup and scalability. The effectiveness and applicability of the proposed algorithm are verified.

• International Journal of Naval Architecture and Ocean Engineering
• /
• 제13권1호
• /
• pp.136-146
• /
• 2021

This paper aims to numerically estimate the dynamic ice load on a conical structure. The Discrete Element Method (DEM) is employed to model the level ice as the assembly of numerous spherical particles. To mimic the realistic fracture mechanism of ice, the parallel bonding method is introduced. Cases with four different ice drifting velocities are considered in time domain. For validation, the statistics of time-varying ice forces and their frequencies obtained by numerical simulations are extensively compared against the physical model-test results. Ice properties are directly adopted from the targeted experimental test set up. The additional parameters for DEM simulations are systematically determined by a numerical three-point bending test. The findings reveal that the numerical simulation estimates the dynamic ice force in a reasonably acceptable range and its results agree well with experimental data.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2011년도 춘계학술발표대회
• /
• pp.250-253
• /
• 2011

Chip manufacture nowadays turned their attention from accelerating uniprocessors to integrating multiple cores on a chip. Moreover desktop graphic hardware is now starting to support general purpose computation. Desktop users are able to use multi-core CPU and GPU as a high performance computing resources these days. However exploiting parallel computing resources are still challenging because of lack of higher programming abstraction for parallel programming. The 2-dimensional discrete cosine transform (2D-DCT) algorithms are most computational intensive part of JPEG encoding. There are many fast 2D-DCT algorithms already studied. We implemented several algorithms and estimated its runtime on multi-core CPU and GPU environments. Experiments show that data parallelism can be fully exploited on CPU and GPU architecture. We expect parallelized DCT bring performance benefit towards its applications such as JPEG and MPEG.

• 천문학회보
• /
• 제44권2호
• /
• pp.71.1-71.1
• /
• 2019

The solar magnetic field comes from the solar interior and is related to various phenomena on the Sun. To understand this process, many studies have been conducted to produce its evolution using a single flux rope. In this study, we are interested in the emergence of two flux ropes and their evolution, which takes longer than the emergence of a single flux rope. To construct it, we develop a flux emergence simulation by applying a parallel computing to reduce a computation time in a wider domain. The original simulation code had been written in Fortran 77. We modify it to a version of Fortran 90 with Message Passing Interface (MPI). The results of the original and new simulation are compared on the NEC SX-Aurora TSUBASA which is a vector engine processor. The parallelized version is faster than running on a single core and it shows a possibility to handle large amounts of calculation. Based on this model, we can construct a complex flux emergence system, such as an evolution of two magnetic flux ropes.

• 예술인문사회 융합 멀티미디어 논문지
• /
• 제4권2호
• /
• pp.369-378
• /
• 2014

The Reed-Muller code is one of the efficient algorithms for multiple bit error correction, however, its high-computation requirement inherent in the decoding process prohibits its use in practical applications. To solve this problem, this paper proposes a graphics processing unit (GPU)-based parallel error control approach using Reed-Muller R(r, m) coding for real-time wireless communication systems. GPU offers a high-throughput parallel computing platform that can achieve the desired high-performance decoding by exploiting massive parallelism inherent in the algorithm. In addition, we compare the performance of the GPU-based approach with the equivalent sequential approach that runs on the traditional CPU. The experimental results indicate that the proposed GPU-based approach exceedingly outperforms the sequential approach in terms of execution time, yielding over 70× speedup.

• 대한기계학회:학술대회논문집
• /
• 대한기계학회 2007년도 춘계학술대회B
• /
• pp.3440-3445
• /
• 2007

For proton exchange membrane fuel cell, it is very important to design the flow channel on separation plate optimally to maximize the current density at same electrochemical reaction surface and reduce the concentration polarization occurred at high current density. In this paper, three dimensional computation model including anode and cathode domain together was developed to examine effects of flow patterns and operation conditions such as humidity and operating temperature on performance of fuel cell. Results show that voltage at counter flow condition is higher than that at coflow condition in parallel and interdigitated flow pattern. And fuel cell with interdigitated flow pattern which has better mass transport by convection flow through gas diffusion layer has higher performance than with parallel flow pattern but its pressure drop is increased such that the trade off between performance and pressure drop should be considered for selection of flow pattern of fuel cell.

• 한국소음진동공학회논문집
• /
• 제15권8호
• /
• pp.962-967
• /
• 2005

The twin-impulse wave discharged from two parallel tubes is investigated to see flow patterns, compared with the single impulse wave. In the present study, the merging phenomena and propagation characteristics of the impulse waves are investigated by experiment and numerical computation. The Harten-Yee's total variation diminishing scheme is used to solve the unsteady, two-dimensional, compressible, Euler equations. The Mach number Ms of incident shock wave is lower than 1.5 and the distance between the tubes is between 1.2 and 4.0. In the shock tube experiment, the twin impulse waves are visualized by a Schlieren optical system in order to validate the computational result. It is shown that on the symmetric axis between two parallel tubes, the peak pressure produced by the twin impulse waves and its location strongly depend upon the tube distance and the incident shock Mach number, Ms. The predicted Schlieren images show a good agreement with the measured twin-impulse wave.

• 한국해안해양공학회지
• /
• 제18권1호
• /
• pp.21-34
• /
• 2006

지진해일 위험재해도의 작성과 재해경감대책 수립을 위해서는 연안역의 상세한 수심 및 지형을 이용한 범람 시뮬레이션이 요구되고 있다. 본 연구에서는 Beowulf 병렬계산을 통해 동해 전 영역에서 정밀산정이 가능한 병렬유한요소모형을 이용하여 1983년 5월 26일 동해안에 내습한 지진해일에 대한 시뮬레이션을 수행하고, 그 계산 결과와 관측치와의 비교결과를 제시한다. 또한, 해안에서의 지진해일고의 통계적 분포에 대해 논하며, 해안에서의 지진해일고의 파고분포가 대수정규분포를 따르는 경향을 제시한다.