• Journal of the Korean Institute of Intelligent Systems
• /
• v.15 no.3
• /
• pp.289-293
• /
• 2005

Pattern classification extracts various types of pattern information expressing objects in the real world and decides their class. The top priority of pattern classification technologies is to improve the performance of classification and, for this, many researches have tried various approaches for the last 40 years. Classification methods used in pattern classification include base classifier based on the probabilistic inference of patterns, decision tree, method based on distance function, neural network and clustering but they are not efficient in analyzing a large amount of multi-dimensional data. Thus, there are active researches on multiple classifier systems, which improve the performance of classification by combining problems using a number of mutually compensatory classifiers. The present study identifies problems in previous researches on multiple SVM classifiers, and proposes BORSE, a model that, based on 1:M policy in order to expand SVM to a multiple class classifier, regards each SVM output as a signal with non-linear pattern, trains the neural network for the pattern and combine the final results of classification performance.

• Bulletin of the Society of Naval Architects of Korea
• /
• v.26 no.1
• /
• pp.39-45
• /
• 1989

Many mechanical systems including ships and/or offshore structures have poor dynamic response characteristics such as undesirable natural frequencies and undesirable mode shapes. It is mandatory to redesign the structure. In this paper a procedure for the dynamic redesign of an undamped structural system is presented. The method which uses a penalty function with a penalty term containing error in equilibrium for a given vibration mode may have a shortcoming. This method includes unconstrained eigenvector degrees of freedom as unknowns. In the work developed here, only constrained mode shape changes are used in the solution procedure, resulting in a reduction of the unnecessary calculations. Among the set of equations which characterizes the redesign of the structural systems, the under constrained problem is discussed here and formulated as an optimization problem, with an optimal criterion such as minimum change or minimum structural weight of the system. Four simple numerical applications illustrate the efficiency of the method. The method can be applied to the vibration problems of ships and/or offshore structures with an implementation of the commercial FE codes.

• Journal of the Society of Naval Architects of Korea
• /
• v.29 no.2
• /
• pp.115-122
• /
• 1992

A sandwich element is a special Hybrid structural form of the composite construction, which is consisted of three main parts : thin, stiff and relatively high density faces separated by a thick, light, and weaker core material. In a sandwich construction, the shear deformation of the faces. Therefore, in the calculation of the bending stiffness, the shear effect should be included. In this paper, the minimum weight is selected as an object function, as the weight critical structures are usually composed of these kind of construction. To obtain the minimum weight of sandwich panel, the principle of minimum potential energy is used and as for the design constraints, the allowable bending stress of face material, the allowable shear stress of core material, the allowable value of panel deflection and the wrinkling stress of faces are adopted, as well as the different boundary conditions. For the engineering purpose of sandwich panel design, the results are tabulated, which are calculated by using the nonlinear optimization technique SUMT.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.16 no.6
• /
• pp.1179-1194
• /
• 1992

QUICKER scheme has several attractive properties. However, under highly convective conditions, it produces overshoots and possibly some oscillations on each side of steps in the dependent variable when the flow is convected at an angle oblique to the grid line. Fortunately, it is possible to modify the QUICKER scheme using non-linear and linear functional relationship. Details of the development of polynomial upwinding scheme are given in this paper, where it is seen that this non-linear scheme has also third order accuracy. This polynomial upwinding scheme is used as the basis for the SHARPER and SMARTER schemes. Another revised scheme was developed by partial modification of QUICKER scheme using CDS and UPWIND schemes(QUICKUP). These revised schemes are tested at the well known bench mark flows, Two-Dimensional Pure Convection Flows in Oblique-Step, Lid Driven Cavity Flows and Buoyancy Driven Cavity Flows. For pure convection oblique step flow test problem, QUICKUP, SMARTER and SHARPER schemes remain absolutely monotonic without overshoot and oscillation. QUICKUP scheme is more accurate than any other scheme in their relative accuracy. In high Reynolds number Lid Driven Cavity Flow, SMARTER and SHARPER schemes retain lower computational cost than QUICKER and QUICKUP schemes, but computed velocity values in the revised schemes produced less predicted values than QUICKER scheme which is strongly effected by overshoot and undershoot values. Also, in Buoyancy Driven Cavity Flow, SMARTER, SHARPER and QUICKUP schemes give acceptable results.

• Journal of Korean Society of Steel Construction
• /
• v.11 no.2 s.39
• /
• pp.153-165
• /
• 1999

In order to trace the lateral-torsional post-bucking behaviors of thin-walled space frames with non-symmetric cross sections, a geometrically non-linear finite element formulation is presented by applying incremental equilibrium equations based on the updated Lagrangian formulation and introducing Vlasov's assumption. The improved displacement field for non-symmetric thin-walled cross sections is introduced based on inclusion of second order terms of finite rotations, and the potential energy corresponding to the semitangential rotations and moments is consistently derived. For finite element analysis, tangent stiffness matrices of thin-walled space frame element are derived by using the Hermition polynomials as shape functions. A co-rotational formulation in order to evaluate the unbalanced loads is presented by separating the rigid body rotations and pure deformations from incremental displacements and evaluating the updated direction cosines and incremental member forces.

• Journal of the Korean Society of Hazard Mitigation
• /
• v.7 no.3
• /
• pp.69-78
• /
• 2007

In this study, a time-dependent aspect of an embankment failure is considered to simulate a flood inundation map and calculate overtopping discharge induced by an embankment failure. A numerical model has been developed by solving the two dimensional nonlinear shallow water equations with a finite volume method on unstructured grids. To analyze a Riemann problem, the HLLC approximate Riemann solver and the Weighted Averaged Flux method are employed by using a TVD limiter and the source term treatment is also employed by using the operator splitting method. Firstly, the numerical model is applied to a dam break problem and a sloping seawall. Obtained numerical results show good agreements with experimental data. Secondly, the model is applied to a flow induced by an embankment failure by assuming that the width and elevation of embankment are varied with time-dependent functions. As a result of the comparison with each numerical overtopping discharge, established flood inundation discharges in the previous studies are overestimated than the result of the present numerical model.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.19 no.6
• /
• pp.72-79
• /
• 2015

An efficient and practical reliability evaluation method is proposed for the coastal structures in this paper. It is capable of evaluating reliability of real complicated coastal structures considering uncertainties in various sources of design parameters, such as wave and current loads, resistance-related design variables including Young's modulus and compressive strength of the reinforced concrete, soil parameters, and boundary conditions. It is developed by intelligently integrating the Latin Hypercube sampling (LHS), Monte Carlo simulation (MCS) and the finite element method (FEM). The LHS-based MCS is used to significantly reduce the computational effort by limiting the number of simulation cycles required for the reliability evaluation. The applicability and efficiency of the proposed method were verified using a caisson-type breakwater structure in the numerical example.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.25 no.5
• /
• pp.437-443
• /
• 2015

This paper presents face recognition based on the fusion of visible image and thermal infrared (IR) texture estimated from the face image in the visible spectrum. The proposed face recognition scheme uses a multi- layer neural network to estimate thermal texture from visible imagery. In the training process, a set of visible and thermal IR image pairs are used to determine the parameters of the neural network to learn a complex mapping from a visible image to its thermal texture in the low-dimensional feature space. The trained neural network estimates the principal components of the thermal texture corresponding to the input visible image. Extensive experiments on face recognition were performed using two popular face recognition algorithms, Eigenfaces and Fisherfaces for NIST/Equinox database for benchmarking. The fusion of visible image and thermal IR texture demonstrated improved face recognition accuracies over conventional face recognition in terms of receiver operating characteristics (ROC) as well as first matching performances.

• Proceedings of the Korea Society for Simulation Conference
• /
• 1993.10a
• /
• pp.12-13
• /
• 1993

컴퓨터를 이용한 모의 실험 방법은 과학 및 공학 분야뿐만 아니라 경제,사회 현상등에도 널리 적용될 수 있는 유용한 도구이다. 그 중에서도 연속체계 모의 실험은 미분 방정식으로 모델링되는 시스템을 대상으로 하는 경우가 맡으며, 이를 위하여 그동안 맡은 연속체계 모의 실험 언어들이 개발되었다. 그러나 그들은 대부분 사용하기가 복잡하여 사용자 편의성을 고려한 모의 실험 언어에 대한 필요성이 증대되었다. 본 연구에서는 사용자에개 최대한 편의성을 제공하는 연속체계 모의 실험 언어인 PCSL (Postech Continuous-system Simulation language)을 개발하였다. PCSL 프로그램은 프로그램 헤더, 상수 정의부, 함수 정의부, 매개 변수 정의부, 초기화 선언부, 모델 정의부, 종료 조건 선언부, 출력 선언부 등으로 나누어 진다. 그리고 출력으로는 계산 결과를 파일에 저장, 흑은 수치로 인쇄하거나 그래프로 그려서 보여준다. PCSL 처리 시스템은 모델 정의부에서 주어진 미분방정식을 해석해서 digital-analog simulation 기법으로 풀 수 있는 형태로 변환하는 번역기와 이렇게 변환된 형태의 미분방정식과 여러 가지 조건들을 고려해서 C 프로그램을 생성해주는 생성기, 생성된 C 프로그램을 실행시켜서 그 결과를 얻는 실행기,그리고 사용자에게 편리한 입출력 방법을 제공하는 사용자 인터페이스로 구성된다. 번역기에서는 모델로 주어진 미분방정식의 종류를 결정한 후에 이들을 digital-analog simulation 기법으로 풀 수 있는 형태로 변환한다. 생성기에서는 번역기의 결과를 받고,프로그램 상의 여러 가지 조건들을 고려해서 C 프로그램을 생성한다. 여기서 생성된 C프로그램은 미분방정식을 포함하는 ‘f.c'와 조건들을 포함하는'main.h', 그리고 digital-analog simulation 기법을 이용하는 모의 실험 알고리즘을 구현한 'main.c'로 구성된다. 그리고 실행기에서는 생성기에서 생성한 C 프로그램을 실행시켜서 결과를 얻는다. 여기에서 필요로 하는 PCSL 프로그램의 내응은 종료 조건 선언부, 출력 선언부 등이다. 마지막으로 사용자 인터페이스는 사용자가 간편하게 PCSL 프로그램을 입력할 수 있게 도와주며 모의 실험 결과를 쉽게 화면상에 보여주기 위한 것이다. 이 때에 사용자가 원하면 계산 결과를 그래프로 그려서 보여주는 기능과 화면에 보이는 결과를 프린터로 출력할 수 있는 기능을 제공한다. 실형 결과로는 먼저 선형 상미분방정식의 예로 mass-damper-spring system, 비선형 상미분방정식의 예로는 van der Pol 방정식, 연립 상미분방정식의 예로는 mixing tank problem 등을 보였으며, 그의 공학에서 일어나는 여러 가지 문제들도 다루었다.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.37 no.3
• /
• pp.55-62
• /
• 2000

This paper presents an analysis of the chaotic behavior in the discrete-time voltage mode chaotic generator fabricated using 0.8${\mu}{\textrm}{m}$ single poly CMOS technology. An approximated empirical equation is extracted from the measurement data of a nonlinear function block. Then the bifurcation diagram is simulated according to input variables and Lyapunov exponent λ which represent a dependence on an initial value is calculated. We show the interrelations among time waveforms, state transition, and power spectra for the state condition of chaotic circuit, such as equilibrium, periodic, and chaotic state. And results of experiments in the chaotic circuit with the $\pm$2.5V power supply and sampling clock frequency of 10KHz are shown and compared with the simulated results.