• Journal of the Korea Society of Computer and Information
• /
• v.20 no.12
• /
• pp.75-82
• /
• 2015

Generally, the optimal solution of assignment problem can be obtained by Hungarian algorithm of two-sided optimization with time complexity $O(n^4)$. This paper suggests one-sided optimal assignment and swap optimization algorithm with time complexity $O(n^2)$ can be achieve the goal of two-sided optimization. This algorithm selects the minimum cost for each row, and reassigns over-assigned to under-assigned cell. Next, that verifies the existence of swap optimization candidates, and swap optimizes with ${\kappa}-opt({\kappa}=2,3)$. For 27 experimental data, the swap-optimization performs only 22% of data, and 78% of data can be get the two-sided optimal result through one-sided optimal result. Also, that can be improves on the solution of best known solution for partial problems.

• KIPS Transactions on Computer and Communication Systems
• /
• v.7 no.3
• /
• pp.67-72
• /
• 2018

Recently, noise reduction in an axial fan producing the small pressure rise and large flow rate, which is one type of turbomachine, is recognized as essential. This study describes the design and optimization techniques of MPI parallel program to simulate the flow-induced noise in the axial fan. In order to simulate the code using 100 million number of grids for flow and 70,000 points for noise sources, we parallelize it using the 2D domain decomposition. However, when it is involved many computing cores, it is getting slower because of MPI communication overhead among nodes, especially for the noise simulation. Thus, it is adopted the one-sided communication to reduce the overhead of MPI communication. Moreover, the allocated memory and communication between cores are optimized, thereby improving 2.97x compared to the original one. Finally, it is achieved 12x and 6x faster using 6,144 and 128 computing cores of KISTI Tachyon2 than using 256 and 16 computing cores for the flow and noise simulations, respectively.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.11 no.18
• /
• pp.71-80
• /
• 1988

Process capability implies the qualitative capability of a process, and it is necessary to specify the process capability by quantification and to evaluate the level with the specified standardizaion. But the process capability index is currently used to evaluate the performance of quality control activity. without considering the characteristics of process structure or the economy of management. Here the researcher would like to redefine process capability and emphasize its index may be used as the measure of managerial assessment and the objective of process quality control in full consideration of the economic aspects of process characteristics. Too little or too much process capability causes a loss of or excess. The proper economic level of process capability varies to each process. The procedure of taking the optimum process capability index is derived from the expected profit function, whereas each method is studied in cases of normal process with one-sided specification and two-sided specification. In addition, the process capability index is presented as a method of quality assurance. And an example is exhibited on wrapping process of 'A' company. The results of this study are summarized at follows. First, though the norminalization of the process capability index is possible by the existing methods, the optimum process capability index can vary to the situation of each process. So, the optimum process capability index which is suggested in this dissertation should be used as the standard to assess process capability. Second, the process capability index can take its effect not just in indicating quality control or managerial records but also in the management of high qualify assurance.

• Journal of Electrical Engineering and Technology
• /
• v.7 no.4
• /
• pp.501-512
• /
• 2012

Social welfare maximization in a double-sided auction market is performed by implementing an aggregation-based particle swarm optimization (CAPSO) algorithm for optimal placement and sizing of one Static Synchronous Series Compensator (SSSC) device. Dallied simulation results (without/with line flow constraints and without/with SSSC) are generated to demonstrate the impact of SSSC on the congestion levels of the modified IEEE 14-bus test system. The proposed CAPSO algorithm employs conventional quadratic smooth and augmented quadratic nonsmooth generator cost curves with sine components to improve the accurate of the model by incorporating the valve loading effects. CAPSO also employs quadratic smooth consumer benefit functions. The proposed approach relies on particle swarm optimization to capture the near-optimal GenCos and DisCos, as well as the location and rating of SSSC while the Newton based load flow solution minimizes the mismatch equations. Simulation results of the proposed CAPSO algorithm are compared to solutions obtained by sequential quadratic programming (SQP) and a recently implemented Fuzzy based genetic algorithm (Fuzzy-GA). The main contributions are inclusion of customer benefit in the congestion management objective function, consideration of nonsmooth generator characteristics and the utilization of a coordinated aggregation-based PSO for locating/sizing of SSSC.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2017.11a
• /
• pp.34-36
• /
• 2017

축류 팬(axial fan)은 팬이 회전하면서 다량의 공기를 불어주는 유체 기계로써 최근 저소음에 대해 중요하게 인식되고 있다. 본 연구는 팬 주위의 소음을 해석하는 MPI 병렬프로그램에 대한 최적화 기법에 대해 다룬다. 이때 수억 개 이상의 격자에서 수만 포인트의 소음을 해석하기 위해서 MPI 프로세스 간 통신이 많이 발생하여 성능이 심각하게 느려지는 현상이 발생한다. 이를 극복하기 위해 MPI 일방향통신을 적용하였다. 뿐만 아니라 통신 및 메모리 최적화 방법을 통해 최대 2.97배 향상시켰다.

• Progress in Medical Physics
• /
• v.30 no.1
• /
• pp.7-13
• /
• 2019

Purpose: To compare the dosimetrical and radiobiological parameters among various volumetric modulated arc therapy (VMAT) techniques using restricted and continuous arc beams for left-sided breast cancer. Materials and Methods: Ten patients with left-sided breast cancer without regional nodes were retrospectively selected and prescribed the dose of 42.6 Gy in 16 fractions on the planning target volume (PTV). For each patient, three plans were generated using the $Eclipse^{TM}$ system (Varian Medical System, Palo Alto, CA) with one partial arc 1pVMAT, two partial arcs 2pVMAT, and two tangential arcs 2tVMAT. All plans were calculated through anisotropic analytic algorithm and photon optimizer with 6 MV photon beam of $VitalBEAM^{TM}$. The same dose objectives for each plan were used to achieve a fair comparison during optimization. Results: For PTV, dosimetrical parameters such as Homogeneity index, conformity index, and conformal number were superior in 2pVMAT than those in both techniques. $V_{95%}$, which indicates PTV coverage, was 91.86%, 96.60%, and 96.65% for 1pVMAT, 2pVMAT, and 2tVMAT, respectively. In most organs at risk (OARs), 2pVMAT significantly reduced the delivered doses compared with the other techniques, excluding the doses to contralateral lung. For the analysis of radiobiological parameters, a significant difference in normal tissue complication probability was observed in ipsilateral lung while no difference was observed in the other OARs. Conclusions: Our study showed that 2pVMAT had better plan quality and normal tissue sparing than 1pVMAT and 2tVMAT but not for all parameters. Therefore, 2pVMAT could be considered the priority choice for the treatment planning for left breast cancer.