• Nuclear Engineering and Technology
• /
• v.50 no.3
• /
• pp.340-355
• /
• 2018

The methods and performance of a fast reactor multigroup cross section (XS) generation code EXUS-F are described that is capable of directly processing Evaluated Nuclear Data File format nuclear data files. RECONR of NJOY is used to generate pointwise XS data, and Doppler broadening is incorporated by the Gauss-Hermite quadrature method. The self-shielding effect is incorporated in the ultrafine group XSs in the resolved and unresolved resonance ranges. Functions to generate scattering transfer matrices and fission spectrum matrices are realized. The extended transport approximation is used in zero-dimensional calculations, whereas the collision probability method and the method of characteristics are used for one-dimensional cylindrical geometry and two-dimensional hexagonal geometry problems, respectively. Verification calculations are performed first for various homogeneous mixtures and cylindrical problems. It is confirmed that the spectrum calculations and the corresponding multigroup XS generations are performed adequately in that the reactivity errors are less than 50 pcm with the McCARD Monte Carlo solutions. The nTRACER core calculations are performed with the EXUS-F-generated 47 group XSs for the two-dimensional Advanced Burner Reactor 1000 benchmark problem. The reactivity error of 160 pcm and the root mean square error of the pin powers of 0.7% indicate that EXUF-F generates properly the broad-group XSs.

• Proceedings of the KIEE Conference
• /
• 2004.07d
• /
• pp.2528-2530
• /
• 2004

Neural network minimization problems are often conditioned and in this contribution way to handle this will be discussed. It is shown that a better conditioned minimization problem can be obtained if the problem is separated with respect to the linear parameters. This will increase the convergence speed of the minimization. One of the most powerful uses of neural networks is in function approximation(curve fitting)[1]. A main characteristic of this solution is that function (f) to be approximated is given not explicitly but implicitly through a set of input-output pairs, named as training set, that can be easily obtained from calibration data of the measurement system. In this context, the usage of Neural Network(NN) techniques for modeling the systems behavior can provide lower interpolation errors when compared with classical methods like polynomial interpolation. This paper solve of single-variable minimization using neural network.

• Structural Engineering and Mechanics
• /
• v.24 no.6
• /
• pp.741-764
• /
• 2006

The time-dependent analysis of prestressed concrete bridges built adopting the incremental launching technique is presented. After summarizing the well known results derived from the elastic analysis, the problem is approached in the visco-elastic domain taking into account the effects consequent to the complex load history affecting the structure. In particular, the effects produced by prestressing applied both in the launching phase and after it and by application of imposed displacements and of delayed restraints during the launching phases are carefully investigated through a refined analytical procedure. The reliability of the proposed algorithm is tested by means of comparisons with reference cases for which exact solutions are known. A case study of general interest is then discussed in detail. This case study demonstrates that a purely elastic approach represents a too crude approximation, which is unable to describe the specific character of the problem.

• Journal of Communications and Networks
• /
• v.6 no.4
• /
• pp.331-342
• /
• 2004

We study the problem of controlling data overhead of mesh-based multicast ad hoc routing protocols by adaptively adding redundancy to the minimal data overhead multicast mesh as required by the network conditions. We show that the computation of the minimal data overhead multicast mesh is NP-complete, and we propose an heuristic approximation algorithm inspired on epidemic algorithms. In addition, we propose a mobility-aware and adaptive mesh construction algorithm based on a probabilistic path selection being able to adapt the reliability of the multicast mesh to the mobility of the network. Our simulation results show that the proposed approach, when implemented into ODMRP, is able to offer similar performance results and a lower average latency while reducing data overhead between 25 to 50% compared to the original ODMRP.

• Journal of the Korean Mathematical Society
• /
• v.44 no.5
• /
• pp.1103-1119
• /
• 2007

We consider multiscale mortar mixed finite element discretizations for slightly compressible Darcy flows in porous media. This paper is an extension of the formulation introduced by Arbogast et al. for the incompressible problem [2]. In this method, flux continuity is imposed via a mortar finite element space on a coarse grid scale, while the equations in the coarse elements (or subdomains) are discretized on a fine grid scale. Optimal fine scale convergence is obtained by an appropriate choice of mortar grid and polynomial degree of approximation. Parallel numerical simulations on some multiscale benchmark problems are given to show the efficiency and effectiveness of the method.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2001.12a
• /
• pp.247-250
• /
• 2001

In this paper, we propose an growing and pruning algorithm to design the adaptive structure of modular wavelet neural network(MWNN) with F-projection and geometric growing criterion. Geometric growing criterion consists of estimated error criterion considering local error and angle criterion which attempts to assign wavelet function that is nearly orthogonal to all other existing wavelet functions. These criteria provide a methodology that a network designer can constructs wavelet neural network according to one's intention. The proposed growing algorithm grows the module and the size of modules. Also, the pruning algorithm eliminates unnecessary node of module or module from constructed MWNN to overcome the problem due to localized characteristic of wavelet neural network which is used to modules of MWNN. We apply the proposed constructing algorithm of the adaptive structure of MWNN to approximation problems of 1-D function and 2-D function, and evaluate the effectiveness of the proposed algorithm.

• KIEAE Journal
• /
• v.9 no.6
• /
• pp.81-87
• /
• 2009

The purpose of the present study is to establish structural noise analyzing method for apartments building floor with structural-acoustic coupling analysis modeling. Noise through floor in the room is recognized as a significant problem with the consequence that noise isolation technique has been studied in the various fields of industry. From among noise factors, resonance sound is the main reason for solid noise of the floor, which is occurred by mechanical vibrations of the acoustic boundary line and the change of velocity. To analyse this phenomenon, numerical computation methods are provided in many fields, In this study, evaluation method for slab is established using finite element method, and a case study for analyzing acoustic phenomenon was suggested. The results show that numerical method, especially F.E.M, has a good approximation to predict noise at floors.

• Bulletin of the Society of Naval Architects of Korea
• /
• v.8 no.1
• /
• pp.29-40
• /
• 1971

In order to study the problem on form factor of hull form, towing test of a full ship model was carried out for different initial trims under both full and half load conditions. The results were fully discussed on the mutual relations among initial trim, mean sinkage and trim change in run, and form factor. There exists optimum initial trim in regard to form factor. Mean sinkage and trim change in run can be expressed in a uninominal approximation in the form of $k_i{\cdot}{F_n}^{2.2}$. The coefficients of this approxmation are related linearly with the initial trim. Form factor changes according to Froude number. It is considered that the trim change in run is a main reason of the fact.

• Korean Journal of Mathematics
• /
• v.32 no.1
• /
• pp.137-162
• /
• 2024

This paper consists of two significant aims. The first aim of this paper is to establish the criteria for the existence of solutions to nonlinear Volterra-Fredholm (V-F) fractional integral equations on [0, L], where 0 < L < ∞. The fractional integral is described here in the sense of the Katugampola fractional integral of order λ > 0 and with the parameter β > 0. The concepts of the fixed point theorem and the measure of noncompactness are used as the main tools to prove the existence of solutions. The second aim of this paper is to introduce a computational method to obtain approximate numerical solutions to the considered problem. This method is based on the Fibonacci wavelets with collocation technique. Besides, the results of the error analysis and discussions of the accuracy of the solutions are also presented. To the best knowledge of the authors, this is the first computational method for this generalized problem to obtain approximate solutions. Finally, two examples are discussed with the computational tables and convergence graphs to interpret the efficiency and applicability of the presented method.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.39 no.1
• /
• pp.16-23
• /
• 2002

In this paper, we propose the growing and pruning algorithm to design the optimal structure of modular wavelet neural network(MWNN) with F-projection and geometric growing criterion. Geometric growing criterion consists of estimated error criterion considering local error and angle criterion which attempts to assign wavelet function that is nearly orthogonal to all other existing wavelet functions. These criteria provide a methodology which a network designer can construct MWNN according to one's intention. The proposed growing algorithm increases in number of module or the size of modules of MWNN. Also, the pruning algorithm eliminates unnecessary node of module or module from constructed MWNN to overcome the problem due to localized characteristic of wavelet neural network which is used to modules of MWNN. We apply the proposed constructing algorithm of the optimal structure of MWNN to approximation problems of 1-D function and 2-D function, and evaluate the effectiveness of the proposed algorithm.