• 한국컴퓨터정보학회논문지
• /
• 제25권9호
• /
• pp.11-20
• /
• 2020

이 논문에서는 HPC 환경에서 컨테이너의 사용 가능성을 탐구하고 적절한 PMI 라이브러리를 선택하는 기준을 실험을 통해 제안한다. HPC 환경에서 컨테이너 기술의 필요성이 대두되고 있지만 Docker 컨테이너 경우 잠재적인 보안 문제가 있어 HPC 환경에 적합하지 않다는 문제가 있다. 이러한 문제를 해결하기 위해 여러 HPC 컨테이너들이 등장하였지만, 서로 다른 특징들 따른 성능 차이가 발생할 수 있다. 그리하여 본 논문에서는 각 HPC 컨테이너 및 Docker 컨테이너 사이의 성능 차이를 NAS Parallel Benchmark 실험을 통해 측정하고 PMI 라이브러리의 종류에 따른 성능 변화를 확인해보았다. 그 결과 HPC 컨테이너와 Docker 컨테이너는 거의 네이티브와 유사한 성능을 보이거나 경우에 따라 오히려 나은 성능이 관측되었다. 또한 PMI 라이브러리 간의 비교에서는 PMIx가 PMI-2에 비해 모든 조건에서 뛰어난 것은 아닌 것으로 관측된 것을 확인할 수 있었다.

• 한국해양공학회:학술대회논문집
• /
• 한국해양공학회 2000년도 추계학술대회 논문집
• /
• pp.79-83
• /
• 2000

A sliding mode fuzzy control (SMFC) with disturbance estimator is applied to design a controller for the third generation benchmark problem on an wind-excited building. A distinctive feature in vibration control of large civil infrastructure is the existence of large disturbances, such as wind, earthquake, and sea wave forces. Those disturbances govern the behavior of the structure, however, they cannot be precisely measured, especially for the case of wind-induced vibration control. Since the structural accelerations are measured only at a limited number of locations without the measurement of the wind forces, the structure of the conventional control may have the feed-back loop only. General structure of the SMFC is composed of a compensation part and a convergent part. The compensation part prevents the system diverge, and the convergent part makes the system converge to the sliding surface. The compensation part uses not only the structural response measurement but also the disturbance measurement, so the SMFC has a feed-back loop and a feed-forward loop. To realize the virtual feed-forward loop for the wind-induced vibration control, disturbance estimation filter is introduced. the structure of the filter is constructed based on an auto regressive model for the stochastic wind force. This filter estimates the wind force at each time instance based on the measured structural responses and the stochastic information of the wind force. For the verification of the proposed algorithm, a numerical simulation is carried out on the benchmark problem of a wind-excited building. The results indicate that the present control algorithm is very efficient for reducing the wind-induced vibration and that the performance indices improve as the filter for wind force estimation is employed.

• 재무관리연구
• /
• 제26권1호
• /
• pp.189-220
• /
• 2009

본 연구는 약형 EMH에 대한 반론으로 제기되어 왔던 모멘텀전략과 반대전략의 우월성에 대하여 검정하였다. 모멘텀전략과 반대전략이 우월한 전략이라면 이는 약형 EMH에 대한 중대한 비판이 된다. 하지만 몬테카를로 시뮬레이션 결과에 따르면 기존의 검정방법은 유의수준왜곡이라는 오류를 갖고 있는 것으로 나타났다. 이에 본 연구는 데이터 스누핑 편의를 해결하는 것으로 알려진 White (2000)의 사실성 체크검정을 이용하여 모멘텀전략과 반대전략의 우월성을 검정하였다. 검정결과는 다음과 같다. 종합주가지수에 대한 정액정기매입전략을 벤치마크 포트폴리오로 정하였을 때 평균수익률을 이용하면 모멘텀전략 중 최선의 전략은 벤치마크 포트폴리오보다 우월한 것으로 나타났다. 하지만 위험을 고려한 성과측정치인 샤프비율을 이용할 경우 모멘텀전략과 반대전략 중 최선의 전략은 우월한 전략이라고 볼 수 없었다. 따라서 위험을 고려한다면 모멘텀전략과 반대전략의 우월성을 근거로 약형 EMH를 기각할 수는 없다.

• Journal of Mechanical Science and Technology
• /
• 제20권12호
• /
• pp.2230-2241
• /
• 2006

This paper presents a sole application of boundary element method to the conjugate heat transfer problem of thermally developing laminar flow in a thick walled pipe when the fluid velocities are fully developed. Due to the coupled mechanism of heat conduction in the solid region and heat convection in the fluid region, two separate solutions in the solid and fluid regions are sought to match the solid-fluid interface continuity condition. In this method, the dual reciprocity boundary element method (DRBEM) with the axial direction marching scheme is used to solve the heat convection problem and the conventional boundary element method (BEM) of axisymmetric model is applied to solve the heat conduction problem. An iterative and numerically stable BEM solution algorithm is presented, which uses the coupled interface conditions explicitly instead of uncoupled conditions. Both the local convective heat transfer coefficient at solid-fluid interface and the local mean fluid temperature are initially guessed and updated as the unknown interface thermal conditions in the iterative solution procedure. Two examples imposing uniform temperature and heat flux boundary conditions are tested in thermally developing region and compared with analytic solutions where available. The benchmark test results are shown to be in good agreement with the analytic solutions for both examples with different boundary conditions.

• Industrial Engineering and Management Systems
• /
• 제10권2호
• /
• pp.134-139
• /
• 2011

The problem of scheduling in permutation flowshops has been extensively investigated by many researchers. Recently, attempts are being made to consider more than one objective simultaneously and develop algorithms to obtain a set of Pareto-optimal solutions. Varadharajan et al. (2005) presented a multi-objective simulated-annealing algorithm (MOSA) for the problem of permutation-flowshop scheduling with the objectives of minimizing the makespan and the total flowtime of jobs. The MOSA uses two initial sequences obtained using heuristics, and seeks to obtain non-dominated solutions through the implementation of a probability function, which probabilistically selects the objective of minimizing either the makespan or the total flowtime of jobs. In this paper, the same problem of heuristically developing non-dominated sequences is considered. We propose an effective heuristics based on simulated annealing (SA), in which the weighted sum of the makespan and the total flowtime is used. The essences of the heuristics are in selecting the initial sequence, setting the weight and generating a solution in the search process. Using a benchmark problem provided by Taillard (1993), which was used in the MOSA, these conditions are extracted in a large-scale experiment. The non-dominated sets obtained from the existing algorithms and the proposed heuristics are compared. It was found that the proposed heuristics drastically improved the performance of finding the non-dominated frontier.

• 대한산업공학회지
• /
• 제33권4호
• /
• pp.469-479
• /
• 2007

This paper focuses on the capacitated vehicle routing problem that determines the routes of vehicles in such a way that each customer must be visited exactly once by one vehicle starting and terminating at the depot while the vehicle capacity and the travel time constraints must be satisfied. The objective is to minimize the total traveling cost. Due to the complexity of the problem, we suggest a tabu search algorithm that combines the features of the existing search heuristics. In particular, our algorithm incorporates the neighborhood reduction method using the proximity information of the Voronoi diagram corresponding to each problem instance. To show the performance of the Voronoi tabu search algorithm suggested in this paper, computational experiments are done on the benchmark problems and the test results are reported.

• 대한산업공학회지
• /
• 제35권2호
• /
• pp.150-159
• /
• 2009

The heterogeneous fleet vehicle routing problem(HVRP) is a variant of the classical vehicle routing problem in which customers are served by a heterogeneous fleet of vehicles with various capacities, fixed costs and variable costs. We propose a new conceptual HVRPCR(HVRP with customer restriction) model including additional customer restrictions in HVRP. In this paper, we develop hybrid particle swarm optimization(HPSO) algorithm with 2-opt and node exchange technique for HVRP. The solution representation is a n-dimensional particle for HVRP with N customers. The decoding method for this representation starts with the transformation of particle into a priority list of customer to enter route and limit of vehicle to serve each customer. The vehicle routes are then constructed based on the customer priority list and limit of vehicle to serve. The proposed algorithm is tested using 8 benchmark problems and it consistently produces high-quality solutions, including new best solutions. The numerical results show that the proposed algorithm is robust and efficient.

• 전자공학회논문지C
• /
• 제35C권3호
• /
• pp.79-89
• /
• 1998

The Error Back-Propagation(EBP) algorithm is widely applied to train a multi-layer perceptron, which is a neural network model frequently used to solve complex problems such as pattern recognition, adaptive control, and global optimization. However, the EBP is basically a gradient descent method, which may get stuck in a local minimum, leading to failure in finding the globally optimal solution. Moreover, a multi-layer perceptron suffers from locking a systematic determination of the network structure appropriate for a given problem. It is usually the case to determine the number of hidden nodes by trial and error. In this paper, we propose a new algorithm to efficiently train a multi-layer perceptron. OUr algorithm uses stochastic perturbation in the weight space to effectively escape from local minima in multi-layer perceptron learning. Stochastic perturbation probabilistically re-initializes weights associated with hidden nodes to escape a local minimum if the probabilistically re-initializes weights associated with hidden nodes to escape a local minimum if the EGP learning gets stuck to it. Addition of new hidden nodes also can be viewed asa special case of stochastic perturbation. Using stochastic perturbation we can solve the local minima problem and the network structure design in a unified way. The results of our experiments with several benchmark test problems including theparity problem, the two-spirals problem, andthe credit-screening data show that our algorithm is very efficient.

• Industrial Engineering and Management Systems
• /
• 제4권1호
• /
• pp.94-101
• /
• 2005

This paper presents the first known application of a meta-heuristic algorithm, variable neighbourhood descent (VND), to the redundancy allocation problem (RAP). The RAP, a well-known NP-hard problem, has been the subject of much prior work, generally in a restricted form where each subsystem must consist of identical components. The newer meta-heuristic methods overcome this limitation and offer a practical way to solve large instances of the relaxed RAP where different components can be used in parallel. The variable neighbourhood descent method has not yet been used in reliability design, yet it is a method that fits perfectly in those combinatorial problems with potential neighbourhood structures, as in the case of the RAP. A variable neighbourhood descent algorithm for the RAP is developed and tested on a set of well-known benchmark problems from the literature. Results on 33 test problems ranging from less to severely constrained conditions show that the variable neighbourhood descent method provides comparable solution quality at a very moderate computational cost in comparison with the best-known heuristics. Results also indicate that the VND method performs with little variability over random number seeds.

• Nuclear Engineering and Technology
• /
• 제28권3호
• /
• pp.311-319
• /
• 1996

We reconstruct the assembly pinwise flux using several types of boundary conditions and confirm that the reconstructed fluxes are the same with the reference flux if the boundary condition is exact. We test EPRI-9R benchmark problem with four boundary conditions, such as Dirichlet boundary condition, Neumann boundary condition, homogeneous mixed boundary condition (albedo type), and inhomogeneous mixed boundary condition. We also test reconstruction of the pinwise flux from nodal values, specifically from the AFEN [1, 2] results. From the nodal flux distribution we obtain surface flux and surface current distributions, which can be used to construct various types of boundary conditions. The result show that the Neumann boundary condition cannot be used for iterative schemes because of its ill-conditioning problem and that the other three boundary conditions give similar accuracy. The Dirichlet boundary condition requires the shortest computing time. The inhomogeneous mixed boundary condition requires only slightly longer computing time than the Dirichlet boundary condition, so that it could also be an alternative. In contrast to the fixed-source type problem resulting from the Dirichlet, Neumann, inhomogeneous mixed boundary conditions, the homogeneous mixed boundary condition constitutes an eigenvalue problem and requires longest computing time among the three (Dirichlet, inhomogeneous mixed, homogeneous mixed) boundary condition problems.