• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 2001.10a
• /
• pp.231-236
• /
• 2001

The goal of multiphase optimization of machine structure is to obtain 1) light weight, 2) statically and dynamically rigid structure. The entire optimization process is carried out in two phases. In the first phase, multiple optimization problem with two objective functions is treated using pareto genetic algorithm. Two objective functions are weight of the structure, and static compliance. In the second phase, maximum receptance is minimized using genetic algorithm. The method is applied to design of quill type machine structure with back column.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2001.05a
• /
• pp.565-570
• /
• 2001

대상 기계구조물의 유한요소 모델로부터 구한 해석결과가 실험결과와 오차를 나타낼 때, 이러한 오차를 줄일 수 있도록 유한요소 모델의 변경이 요구된다. 유한요소 모델개선은 이러한 역문제(Inverse Problem)를 다루는 체계적인 접근법이다. 일반적으로 유한요소 모델에서 변경할 수 있는 매개변수의 개수는 실험결과의 개수보다 많으므로 실험결과와 일치되는 개선된 유한요소 모델은 무한하다고 할 수 있다. 그러나, 개선된 유한요소 모델이 물리적 타당성을 갖도록 매개변수의 변경량에 제한을 주면 일반적으로 초기 유한요소 모델에 비해 실험결과와의 오차가 개선된 근사해만 존재하게 된다. 따라서, 모델개선 과정을 통해 구한 개선된 모델은 오차의 평가기준 또는 목적함수에 따라 정해진 다양한 근사해 중 하나이다. 기존의 모델개선 방법에서는 단 하나의 오차 평가기준 또는 목적함수를 사용하고 이를 최소화 하는 모델을 구한다. 개선된 모델을 구하기 이전에는 사용된 평가기준이 타당한지 검토할 수 없으므로 대부분의 경우, 시행착오법으로 목적함수를 설정하게 된다. 본 논문에서는 다목적 최적화 기법을 이용한 오차 평가기준을 소개하고 이를 하드디스크커버 유한요소 모델개선에 응용한다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.17 no.7
• /
• pp.869-874
• /
• 2007

In searching for solutions to multiobjective optimization problem, we find that there is no single optimal solution but rather a set of solutions known as 'Pareto optimal set'. To find approximation of ideal pareto optimal set, search capability of diverse individuals at population space can determine the performance of evolutionary algorithms. This paper propose the method to maintain population diversify and to find non-dominated alternatives in Game model based Co-Evolutionary Algorithm.

• Annual Conference on Human and Language Technology
• /
• 2016.10a
• /
• pp.333-336
• /
• 2016

본 연구에서는 개체명 인식의 성능을 향상시키기 위해, 가중 투표 방법을 이용하여 개체명 인식 모델을 앙상블 하는 방법을 제안한다. 각 모델은 Conditional Random Fields의 변형 알고리즘을 사용하여 학습하고, 모델들의 가중치는 다목적 함수 최적화 기법인 NSGA-II 알고리즘으로 학습한다. 실험 결과 제안 시스템은 $F_1Score$기준으로 87.62%의 성능을 보여, 단독 모델 중 가장 높은 성능을 보인 방법보다 2.15%p 성능이 향상되었다.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2002.05a
• /
• pp.16-22
• /
• 2002

STEP Method(STEM)는 의사결정자와 최적화 모형간의 상호작용을 통하여 문제를 해결하는 다목적 최적화 기법이다 STEM은 최적화 모형을 계산하는 계산 단계와 계산 단계에서 도출된 결과에 대하여 의사결절자의 선호도 정보를 모형에 반영하는 의사결정 단계로 구성되어 있다. STEM의 두 단계에서는 의사결정자의 선호도 정보가 불확실성을 포함한 경우를 적절히 고려하지 못하고 있다. 본 연구에서는 퍼지모델링 기법을 사용하여 STEM의 문제점을 보완한 기법을 제안하고자 한다.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.31 no.5
• /
• pp.243-250
• /
• 2018

In this paper, design technique using genetic algorithms(GA) for design optimization of composite leaf springs is presented here. After the initial design has been validated by the car plate spring as a finite element model, the genetic algorithm suggests the process of optimizing the number of layers of composite materials and their angles. Through optimization process, the weight reduction process of leaf springs and the number of repetitions are compared to the existing algorithm results. The safety margin is calculated by organizing a finite element model to verify the integrity of the structure by applying an additive sequence optimized through the genetic algorithm to the structure. When GA is applied, layer thickness and layer angle of complex leaf springs have been obtained, which contributes to the achievement of minimum weight with appropriate strength and stiffness. A reduction of 65.6% original weight is reached when a leaf steel spring is replaced with a leaf composite spring under identical requirement of design parameters and optimization.

• Journal of Korea Water Resources Association
• /
• v.40 no.9
• /
• pp.687-696
• /
• 2007

Preference ordering approach is applied to optimize the parameters of Tank model using multi-objective genetic algorithm (MOGA). As more than three multi-objective functions are used in MOGA, too many non-dominated optimal solutions would be obtained thus the stakeholder hardly find the best optimal solution. In order to overcome this shortcomings of MOGA, preference ordering method is employed. The number of multi-objective functions in this study is 4 and a single Pareto-optimal solution, which is 2nd order efficiency and 3 degrees preference ordering, is chosen as the most preferred optimal solution. The comparison results among those from Powell method and SGA (simple genetic algorithm), which are single-objective function optimization, and NSGA-II, multi-objective optimization, show that the result from NSGA-II could be reasonalby accepted since the performance of NSGA-II is not deteriorated even though it is applied to the verification period which is totally different from the calibration period for parameter estimation.

• Korean Journal of Air-Conditioning and Refrigeration Engineering
• /
• v.17 no.10
• /
• pp.938-946
• /
• 2005

Shape optimization of internally finned circular tube has been peformed for periodically fully developed turbulent flow and heat transfer. The physical domain considered in this study is very complicated due to periodic boundary conditions both streamwise and circumferential directions. Therefore, Pareto frontier sets of a heat exchanger can be acquired by coupling the CFD and the multi-objective genetic algorithm, which is a global optimization technique. The optimal values of fin widths $(d_1,\;d_2)$ and fin height (H) are numerically obtained by minimizing the pressure loss and maximizing the heat transfer rate within ranges of $d_1=0.2\sim1.5\;mm,\;d_2=0.2\sun1.5\;mm,\;and\;H=0.2\sim1.5\;mm$. The optimal values of the design variables are acquired after the fifth generation and also compared to those of a local optimization algorithm for the same geometry and conditions.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2000.04a
• /
• pp.53-56
• /
• 2000

The problem ( Ρ ) of optimizing a linear function d$\^$T/x over the set of efficient set for a multiple objective linear program ( Μ ) is difficult because the efficient set is nonconvex. There some interesting properties between the objective linear vector ｄ and the matrix of multiple objectives C and those properties lead us to establish criteria to solve ( Ρ ) with a linear program. In this paper we investigate a system of the linear equations C$\^$T/${\alpha}$=d and construct two linearly independent positive vectors ${\mu}$, ν such that ${\alpha}$=${\mu}$-ν. From those vectors ${\mu}$, ν, solving an weighted sum linear program for finding an efficient extreme point for the ( Μ ) is a way to get an optimal solution ( Ρ ). Therefore our theory gives an easy way of solving nonconvex program ( Ρ ) with a weighted sum linear program.

• Journal of Korea Water Resources Association
• /
• v.40 no.9
• /
• pp.677-685
• /
• 2007

The objective of this study is to evaluate the applicability of multi-objective genetic algorithm(MOGA) in order to calibrate the parameters of conceptual rainfall-runoff model, Tank model. NSGA-II, one of the most imitating MOGA implementations, is combined with Tank model and four multi-objective functions such as to minimize volume error, root mean square error (RMSE), high flow RMSE, and low flow RMSE are used. When NSGA-II is employed with more than three multi-objective functions, a number of Pareto-optimal solutions usually becomes too large. Therefore, selecting several preferred Pareto-optimal solutions is essential for stakeholder, and preference-ordering approach is used in this study for the sake of getting the best preferred Pareto-optimal solutions. Sensitivity analysis is performed to examine the effect of initial genetic parameters, which are generation number and Population size, to the performance of NSGA-II for searching the proper paramters for Tank model, and the result suggests that the generation number is 900 and the population size is 1000 for this study.