• Journal of the Korean Data and Information Science Society
• /
• v.27 no.2
• /
• pp.381-394
• /
• 2016

Graphical model represents conditional dependencies between variables as a graph with nodes and edges. It is widely used in various fields including physics, economics, and biology to describe complex association. Conditional dependencies can be estimated from a inverse covariance matrix, where zero off-diagonal elements denote conditional independence of corresponding variables. This paper proposes a efficient BCDR (block coordinate descent with random permutation) algorithm using graphics processing units and random permutation for the CONCORD (convex correlation selection method) based on the BCD (block coordinate descent) algorithm, which estimates a inverse covariance matrix based on pseudo-likelihood. We conduct numerical studies for two network structures to demonstrate the efficiency of the proposed algorithm for the CONCORD in terms of computation times.

• Journal of IKEEE
• /
• v.19 no.1
• /
• pp.33-40
• /
• 2015

Inverse problem in electrical impedance tomography (EIT) is highly ill-posed therefore prior information is used to mitigate the ill-posedness. Regularization methods are often adopted in solving EIT inverse problem to have satisfactory reconstruction performance. In solving the EIT inverse problem, iterative Gauss-Newton method is generally used due to its accuracy and fast convergence. However, its performance is still suboptimal and mainly depends on the selection of regularization parameter. Although, there are few methods available to determine the regularization parameter such as L-curve method they are sometimes not applicable for all cases. Moreover, regularization parameter is a scalar and it is fixed during iteration process. Therefore, in this paper, a novel method is used to determine the regularization parameter to improve reconstruction performance. Conductivity norm is calculated at each iteration step and it used to obtain the regularization parameter which is a diagonal matrix in this case. The proposed method is applied to human thorax imaging and the reconstruction performance is compared with traditional methods. From numerical results, improved performance of proposed method is seen as compared to conventional methods.

• Composites Research
• /
• v.13 no.4
• /
• pp.54-66
• /
• 2000

Interfacial and microfailure properties of carbon fiber/epoxy matrix composites were evaluated using both tensile fragmentation and compressive Broutman tests with an aid of acoustic emission (AE) monitoring. A polymeric maleic anhydride coupling agent and a monomeric amino-silane coupling agent were used via the electrodeposition (ED) and the dipping applications, respectively. Both coupling agents exhibited significant improvements in interfacial shear strength (IFSS) compared to the untreated case under tensile and compressive tests. The typical microfailure modes including fiber break of cone-shape, matrix cracking, and partial interlayer failure were observed during tensile test, whereas the diagonal slippage in fiber ends was observed under compressive test. For both loading types, fiber breaks occurred around just before and after yielding point. In both the untreated and treated cases AE amplitudes were separately distributed for the tensile testing, whereas they were closely distributed for the compressive tests. It is because of the difference in failure energies of carbon fiber between tensile and compressive loading. The maximum AE voltage for the waveform of carbon or basalt fiber breakages under tensile tests exhibited much larger than those under compressive tests, which can provide the difference in the failure energy of the individual failure processes.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.12 no.1
• /
• pp.92-98
• /
• 2001

When analyzing the scatting problem of multi-layered planar structures using closed-form Green's function, one of the main difficulties is that the numerical integrations for the evaluation of diagonal matrix elements converge slowly and are not so stable. Accordingly, even when the integration fur the singularity of type $e^{-jkr}/{\gamma}$, corresponding to the source dipole itself, is performed using such a method, this difficulty persists in the integration corresponding to the finite number of complex images. In order to resolve this difficulty, a new technique based upon the Gaussian quadrature in polar coordinates for the evaluation of the two-dimensional generalized exponential integral is presented. Stability of the algorithm and convergence is discussed. Performance is demonstrated for the example of a microstrip patch antenna.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.32 no.1A
• /
• pp.51-56
• /
• 2007

Unitary matrix modulation (UMM) is investigated in multiple antennas system that is called unitary space-time modulation (USTM). When we consider only diagonal components of UMM with splitting over the coherence bandwidth, the system can obtain frequency diversity in a single antenna system. In Rayleigh fading channel, we compared USTM/OFDM with 2Tx-antenna with UMM-S/OFDM with 1Tx-antenna, conventional multi-level OFDM and UMM-S/OFDM with changing bit rate per Hz. Also, we analyzed SNR for between multi-level modulation STBC/OFDM and multi-level modulation UMM-S/OFDM with changing transmission rate. When it was adaptive multi-level modulation to improve SNR, we did the UMM-S/OFDM system performance analysis of N-ary PSK and M-ary QAM.

• Journal of Ocean Engineering and Technology
• /
• v.11 no.2
• /
• pp.100-106
• /
• 1997

Viscous flow around an actual ship is calculated by an use of RANS(Reynolds-averaged Navier-Stokes) solver. Reynolds stress is modelled by using k-$\varepsilon$ turbulence model and the law of wall is applied near the body. Body fitted coordinates are introduced for the treatment of the complex boundary of the ship hull form. The transformed equations in the computational domain are numerically solved by an employment of FVM(Finite Volume Method). SIMPLE(Semi-Implcit Pressure Linked Equation) method is adopted in the calculation of pressure and the solution of the disssssssscretized equation is obtained by the line-by-line method with the use of TDMA(Tri-Diagonal Matrix Algorithme). The subject ship model of actual calculation is 4,410 TEU class container carrier. For 4 geosim models the calculated viscous resistancce values are compared with the model test results and analyzed on their componentss. The resistance performance of an actual ship is predicted very resonably, so this mothod may be utilized as a design tool of hull form.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.4 no.5
• /
• pp.16-25
• /
• 1996

In the present work a rigid-plastic finite element formulation using dynamic explicit time integration scheme is proposed for numerical analysis of auto-body panel stamping processes. The rigid-plastic finite element method based on membrane elements has long been employed as a useful numerical technique for the analysis of sheet metal forming because of its time effectiveness. A damping scheme is proposed in order to achieve a stable solution procedure in dynamic sheet forming problems. In order to improve the drawbacks of the conventional membrane elements, BEAM(abbreviated from Bending Energy Augmented Membrane) elements are employed. Rotational damping and spring about the drilling direction are introduced to prevent a zero energy mode. The lumping scheme is employed for the diagonal mass matrix and linearizing dynamic formulation. A contact scheme is developed by combining the skew boundary condition and the direct trial-and-error method. Computations are carried out for analysis of complicated auto-body panel stamping processes such as forming of an oilpan, a fuel tank and a front fender. The numerical results of explicit analysis are compared with the implicit results with good agreements and it is shown that the explicit scheme requires much shorter computational time, especially when the problem becomes more complicated. It is thus shown that the proposed dynamic explicit rigid-plastic finite element method enables an effective computation for complicated autobody panel stamping processes.

• Proceedings of the Korea Committee for Ocean Resources and Engineering Conference
• /
• 2002.05a
• /
• pp.161-166
• /
• 2002

Low-tension cables have been increasingly used in recent years due to deep-sea developments and the advent of synthetic cables. In the case of low-tension cables, large displacements may happen due to relatively small restoring forces of tension and thus the effects of fluid and geometric non-linearities become predominant. In this study, three-dimensional (3-D) dynamic behavior of a towed low-tension cable with non-uniform characteristics is numerically analyzed by considering fluid and geometric non-linearities and bending stiffness. A Fortran program is developed by employing a finite difference method. In the algorithm, an implicit time integration and Newton-Raphson iteration are adopted. For the calculation of huge size of matrices, block tri-diagonal matrix method is applied, which is much faster than the well-known Gauss-Jordan method in two point boundary value problems. Some case studies are carried out and the results of numerical simulations are compared with a in-house program of WHOI Cable with good agreements.

• Journal of Mechanical Science and Technology
• /
• v.16 no.2
• /
• pp.203-210
• /
• 2002

For effective response surface modeling during sequential approximate optimization (SAO), the normalized and the augmented D-optimality criteria are presented. The normalized D-optimality criterion uses the normalized Fisher information matrix by its diagonal terms in order to obtain a balance among the linear-order and higher-order terms. Then, it is augmented to directly include other experimental designs or the pre-sampled designs. This augmentation enables the trust region managed sequential approximate optimization to directly use the pre-sampled designs in the overlapped trust regions in constructing the new response surface models. In order to show the effectiveness of the normalized and the augmented D-optimality criteria, following two comparisons are performed. First, the information surface of the normalized D-optimal design is compared with those of the original D-optimal design. Second, a trust-region managed sequential approximate optimizer having three D-optimal designs is developed and three design problems are solved. These comparisons show that the normalized D-optimal design gives more rotatable designs than the original D-optimal design, and the augmented D-optimal design can reduce the number of analyses by 30% - 40% than the original D-optimal design.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.11 no.5
• /
• pp.755-761
• /
• 1987

Two-Dimensional incompressible laminar boundary layer with the reversed flow region is computed using the parially parabolized Navier-Stokes equations in primitive variables. The velocities and the pressure are explicity coupled in the difference equation and the resulting penta-diagonal matrix equations are solved by a streamwise marching technique. The test calculations for the trailing edge region of a finite flat plate and Howarth's linearly retarding flows demonstrate that the method is accurate, efficient and capable of predicting the reversed flow region.