• Journal of Advanced Marine Engineering and Technology
• /
• v.22 no.5
• /
• pp.693-701
• /
• 1998

This study presents result of comparison between two other numerical method method of char-acteristics(MOC) and Lax-Wendroff method(LWM) applied at wave action analysis of intake and exhaust pipe in terms of calculated pressure velocity and emitted noise in the time and the fre-quency domain by means of fast Fourier transform analysis. Particularly FCT(Flux Corrected Transport)scheme is appended to LWM to protest unaceptable overshoots occurring near discon-tinuity. The final conclusion of this study is that MOC should be replaced by a second order finite difference approach because of larger contributions due to high frequency components than the results from the method of characteristics. Clear benefits we can get by change are faster calcula-tion higher accuracy conservation of mass and consistent calculation method.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.21 no.2
• /
• pp.186-193
• /
• 2011

The NDIF method(non-dimensional dynamic influence function method) for free vibration analysis of arbitrarily shaped plates with the simply supported edge is newly developed in the paper. In order to extract the system matrix that gives the natural frequencies and natural modes of the plate of interest, the difficulty of measuring higher differential terms involved in the simply supported boundary condition is successfully overcome. Finally, the excellence of the characteristics of convergence and accuracy of the proposed method is shown through two verification examples, which indicate that natural frequencies and natural modes obtained by the proposed method are very accurate and swiftly converged even though a small number of nodes are used compared with FEM.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.4 no.4
• /
• pp.72-79
• /
• 1996

The recent trend of higher output engines with more auxiliary parts is resulting in greater heat generation in the engine compartment. In order to maximize the heat dissipation and eliminate the inefficient flow in the engine compartment, it is necessary to understand the flow field under the hood. In this respect, experimental study as well as numerical analysis should be conducted. The automated measuring system was constructed to obtain three dimensional mean flow data with high accuracy. The measurements have been made on a vehicle with a steady incoming air flow. The result shows that there exists a high degree of non-uniformity in the mean flow velocity at the front of radiator.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.40 no.2
• /
• pp.101-107
• /
• 2012

In this study, Constraint Force Equation Methodology (CFEM) is used to develop a multi-body dynamic analysis program with constraints. Seven constraint models are implemented to analyze constraint motions of multiple bodies. The augmented equations with the constraints are solved with the 4th order Runge-Kutta method for higher degree of accuracy. The analysis code is verified by comparing the analysis results of the motion of bodies with various constraints to published results.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.3 no.3
• /
• pp.309-330
• /
• 2009

In this paper, we introduce the architecture of Genetic Algorithm(GA) based Feed-forward Polynomial Neural Networks(PNNs) and discuss a comprehensive design methodology. A conventional PNN consists of Polynomial Neurons, or nodes, located in several layers through a network growth process. In order to generate structurally optimized PNNs, a GA-based design procedure for each layer of the PNN leads to the selection of preferred nodes(PNs) with optimal parameters available within the PNN. To evaluate the performance of the GA-based PNN, experiments are done on a model by applying Medical Imaging System(MIS) data to a multi-variable software process. A comparative analysis shows that the proposed GA-based PNN is modeled with higher accuracy and more superb predictive capability than previously presented intelligent models.

• Journal of the Korea Society of Computer and Information
• /
• v.23 no.6
• /
• pp.1-7
• /
• 2018

A multi-label classification is to find multiple labels associated with the input pattern. Multi-label classification can be achieved by extending conventional single-label classification. Common extension techniques are known as Binary relevance, Label powerset, and Classifier chains. However, most of the extended multi-label naive bayes classifier has not been able to accurately estimate posterior probabilities because it does not reflect the label dependency. And the remaining extended multi-label naive bayes classifier has a problem that it is unstable to estimate posterior probability according to the label selection order. To estimate posterior probability well, we propose a new posterior probability estimation method that reflects the probability between all labels and labels efficiently. The proposed method reflects the correlation between labels. And we have confirmed through experiments that the extended multi-label naive bayes classifier using the proposed method has higher accuracy then the existing multi-label naive bayes classifiers.

• Structural Engineering and Mechanics
• /
• v.43 no.2
• /
• pp.253-271
• /
• 2012

To analyze the bending and transverse shear effects of laminated composite plates, a thirteen nodes triangular element will be presented. The suggested formulations consider a parabolic variation of the transverse shear strains through the thickness. As a result, there is no need to use shear correction coefficients in computing the shear stresses. The proposed element can model both thin and thick plates without any problems, such as shear locking and spurious modes. Moreover, the effectiveness of $w_{,n}$, as an independent degree of freedom, is concluded by the present study. To perform the accuracy tests, several examples will be solved. Numerical results for the orthotropic materials with different boundary conditions, shapes, number of layers, thickness ratios and fiber orientations will be presented. The suggested element calculates the deflections and stresses more accurate than those available in the literature.

• Journal of Power System Engineering
• /
• v.4 no.1
• /
• pp.68-73
• /
• 2000

A mixed degree finite element solutions using hierarchical elements are investigated for convergences on a 2-D simple cases. Elements are generated block by block and each block is assigned an arbitrary solution degree. The numerical study showed that a well constructed blocks can increase the convergence and accuracy of finite element solutions. Also, it has been found that for higher order elements, the convergence trends can be deteriorated for smaller mesh sizes. A procedure for a variable fixed boundary condition has been included.

• Journal of the Society of Naval Architects of Korea
• /
• v.41 no.4
• /
• pp.17-29
• /
• 2004

Internal waves are computed using a ghost fluid method on an unstructured grid. Discontinuities in density and dynamic pressure are captured in one cell without smearing or oscillations along a multimaterial interface. A time-accurate incompressible Navier-Stokes/Euler solver is developed based on a three-point backward difference formula for the physical time marching. Artificial compressibility is introduced with respect to pseudotime and an implicit method is used for the pseudotime iteration. To track evolution of an interface, a level set function is coupled with the governing equations. Roe's flux difference splitting method is used to calculate numerical fluxes of the coupled equations. To get higher order accuracy, dependent variables are reconstructed based on gradients which are calculated using Gauss theorem. For each edge crossing an interface, dynamic pressure is assigned for a ghost node to enforce the continuity of total pressure along the interface. Solitary internal waves are computed and the results are compared with other computational and experimental results.

• ETRI Journal
• /
• v.40 no.5
• /
• pp.604-612
• /
• 2018

To solve the problem of unbalanced loads and the short network lifetime of heterogeneous wireless sensor networks, this paper proposes a node-selection algorithm based on energy balance and dynamic adjustment. The spacing and energy of the nodes are calculated according to the proximity to the network nodes and the characteristics of the link structure. The direction factor and the energy-adjustment factor are introduced to optimize the node-selection probability in order to realize the dynamic selection of network nodes. On this basis, the target path is selected by the relevance of the nodes, and nodes with insufficient energy values are excluded in real time by the establishment of the node-selection mechanism, which guarantees the normal operation of the network and a balanced energy consumption. Simulation results show that this algorithm can effectively extend the network lifetime, and it has better stability, higher accuracy, and an enhanced data-receiving rate in sufficient time.