• ETRI Journal
• /
• v.5 no.3
• /
• pp.3-8
• /
• 1983

In the area of computer network planning, a location-allocation-size problem is involved. Since multi-facility location-allocation-size problems are very complex in formulating a mathematical model, it is a usual practise to adopt alternative approaches, which give no optimal results, instead of the optimal solution by mathematical approach. In this article, however, an attempt is made to formulate a mathematical model for the decision making problem of computer network design.

• Computational Structural Engineering
• /
• v.10 no.4
• /
• pp.349-358
• /
• 1997

The present paper describes an investigation into the application of the genetic algorithm (GA) in the optimum design of composite laminated structure. Stochastic processes generate an initial population of designs and then apply principles of natural selection/survival of the fittest to improve the designs. The five test functions are used to verify the robustness and reliability of the GA, and as a numerical example, minimum weight of a cantilever composite laminated beam with a mix of continuous, integer and discrete design variables is obtained by using the GA with exterior penalty function method. The design problem has constraints on strength, displacements, and natural frequencies, and is formulated to a multidimensional nonlinear form. From the results, it is found that the GA search technique is very effective to find the good optimum solution as well as has higher robustness.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.10C
• /
• pp.1402-1413
• /
• 2004

In non-rigid image registration, the Jacobian determinant of the estimated deformation should be positive everywhere since physical deformations are always invertible. We propose a constrained optimization technique at ensures the positiveness of Jacobian determinant for cubic B-spline based deformation. We derived sufficient conditions for positive Jacobian determinant by bounding the differences of consecutive coefficients. The parameter set that satisfies the conditions is convex; it is the intersection of simple half spaces. We solve the optimization problem using a gradient projection method with Dykstra's cyclic projection algorithm. Analytical results, simulations and experimental results with inhale/exhale CT images with comparison to other methods are presented.

• Wind and Structures
• /
• v.1 no.1
• /
• pp.111-125
• /
• 1998

This paper deals with the development of variable-node element and its application to the adaptive h-version mesh refinement-recovery for the incompressible viscous flow analysis. The element which has variable mid-side nodes can be used in generating the transition zone between the refined and unrefined element and efficiently used for the construction of a refined mesh without generating distorted elements. A modified Guassian quadrature is needed to evaluate the element matrices due to the discontinuity of derivatives of the shape functions used for the element. The penalty function method which can reduce the number of the independent variables is adopted for the purpose of computational efficiency and the selective reduced integration is carried out for the convection and pressure terms to preserve the stability of solution. For the economical analysis of transient problems in which the locations to be refined are changed in accordance with the dynamic distribution of velocity gradient, not only the mesh refinement but also the mesh recovery is needed. The numerical examples show that the optimal mesh for the finite element analysis of a wind around the structures can be obtained automatically by the proposed scheme.

• Nuclear Engineering and Technology
• /
• v.56 no.1
• /
• pp.265-274
• /
• 2024

The current research focuses on the development of a numerical approach to forecast strongly subcooled flow boiling of FC-72 as the refrigerant in various vertical minichannel shapes for high-heat-flux cooling applications. The simulations are carried out using the Volume of Fluid method with the Lee phase change model, which revealed some inherent flaws in multiphase flows that are primarily due to an insufficient interpretation of shearlift force on bubbles and conjugate heat transfer against the walls. A user-defined function (UDF) is used to provide specific information about this noticeable effect. The influence of shape and the inlet mass fluxes on the flow patterns, heat transfer, and pressure drop characteristics are discussed. The computational results are validated with experimental measurements, where excellent agreements are found that prove the efficiency of the present numerical model. The findings demonstrate that the heat transfer coefficient decreases as the mass flux increases and that the constriction design improves the thermal performance by 24.68% and 10.45% compared to the straight and expansion shapes, respectively. The periodic constriction sections ensure good mixing between the core and near-wall layers. In addition, a slight pressure drop penalty versus the thermal transfer benefits for the two configurations proposed is reported.

• Journal of Digital Contents Society
• /
• v.16 no.4
• /
• pp.605-613
• /
• 2015

This paper addresses the problem to make sparse the projection matrix in pattern recognition method. Recently, the size of computer program is often restricted in embedded systems. It is very often that developed programs include some constant data. For example, many pattern recognition programs use the projection matrix for dimension reduction. To improve the recognition performance, very high dimensional feature vectors are often extracted. In this case, the projection matrix can be very big. Recently, RSR(roated sparse regression) method[1] was proposed. This method has been proved one of the best algorithm that obtains the sparse matrix. We propose three methods to improve the RSR; outlier removal, sampling and elastic net RSR(E-RSR) in which the penalty term in RSR optimization function is replaced by that of the elastic net regression. The experimental results show that the proposed methods are very effective and improve the sparsity rate dramatically without sacrificing the recognition rate compared to the original RSR method.

• Journal of the Korean Magnetics Society
• /
• v.24 no.5
• /
• pp.160-164
• /
• 2014

In this paper, an efficient robust design utilizing an enhanced Taguchi method is proposed to reduce cogging torque of a BLDC motor in the presence of design uncertainty. To overcome defects of the conventional Taguchi method in dealing with a generalized robust design problem, a penalty function and an optimal level searching technique are newly introduced. In order to verify the proposed method, a 5 kW, rated speed of 2,300 rpm, rated torque of 20 Nm BLDC motor for driving electric vehicles is optimized. Then, the robust design is compared with conceptual and deterministic ones in terms of the cogging torque, rated torque and torque ripple.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.10 no.3
• /
• pp.175-185
• /
• 2006

This study performed the optimum design of balanced and unbalanced span length bridges with many variable Girder types by using the optimum design program to minimize the cost for PSC box girder bridge of the full staging method. The objective of this study is to present tendon's application direction about complicated construction hereafter by studying about optimum tendon arrangement that is worked in each variable Girder type. This program used SUMT procedure and Kavlie's extended penalty function to allow infeasible design points in the process. Powell's direct method was used in searching design points and Gradient Approximate Method was used to reduce design hours.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.14 no.4
• /
• pp.495-503
• /
• 2001

In a functionally graded materials(FGM), two constituent material particles are mixed up according to a specific volume fraction distribution so that its thermoelastic behavior is definitely characterized by such a material composition distribution. Therefore, the designer should determine the most suitable volume fraction distribution in order to design a FGM that optimally meets the desired performance against the given constraints. In this paper, we address a numerical optimization procedure, with employing interior penalty function method(IPFM) and FDM, for optimizing 2D volume fractions of heat-resisting FGMs composed of metal and ceramic. We discretize a FGM domain into finite number of homogenized rectangular cells of single design variable in order for the optimization efficiency. However, after the optimization process, we interpolate the discontinuous volume fraction with globally continuous bilinear function in order to enforce the continuity of volume fraction distributions.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.17 no.3
• /
• pp.309-318
• /
• 2004

This research automatically designed psc-box girder bridges by using an optimum design program and applied the results to the various types of bridges to verify if common facts used in steel bridges or concrete bridges can be applied to PSC bridges. Namely, it investigated appropriate unequal span length division by comparing with bridge of unequal and equal span length division, and verified the influence of the load factors which are changed by time or specification applying the results to various types of bridge. and it applied reinforced concrete bridge and steel bridge's variable girder depth which is slender and effective to save material costs to PSC box girder bridges. Technical solution of optimum design program used SUMT procedure, and Kavlie's extended penalty function to allow infeasible design points in the process. Powell's direct method was used for searching design points and a gradient's approximate method was used to reduce the design time.