• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.4 no.2
• /
• pp.397-406
• /
• 2000

Genetic algorithm is used as a means of search, optimization md machine learning, its structure is simple but it is applied to various areas. And it is about an active and effective controller which can flexibly prepare for changeable circumstances. For this study, research about an action base system evolving by itself is also being considered. There is to have a problem that depended entirely on heuristic knowledge of expert forming membership function and control rule for fuzzy controller design. In this paper, for forming the fuzzy control to perform self-organization, we tuned the membership function to the most optimal using a genetic algorithm(GA) and improved the control efficiency by the self-correction and generation of control rules.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.46 no.5
• /
• pp.70-77
• /
• 2009

In an RFID network consisting of a single reader and many tags, a framed and slotted ALOHA, which provides a number of slots for the tags to respond, was introduced for arbitrating a collision among tags' responses. In a framed and slotted ALOHA, the number of slots in each frame should be optimized to attain the maximal efficiency in tag cognizance. While such an optimization necessitates the knowledge about the number of tags, the reader hardly knows it. In this paper, we propose a tag cognizance scheme based on framed and slotted ALOHA, which is characterized by directly taking a Bayes action on the number of slots without estimating the number of tags separately. Specifically, a Bayes action is yielded by solving a decision problem which incorporates the prior distribution the number of tags, the observation on the number of slots in which no tag responds and the loss function reflecting the cognizance rate. Also, a Bayes action in each frame is supported by an evolution of prior distribution for the number of tags. From the simulation results, we observe that the pair of evolving prior distribution and Bayes action forms a robust scheme which attains a certain level of cognizance rate in spite of a high discrepancy between the Due and initially believed numbers of tags. Also, the proposed scheme is confirmed to be able to achieve higher cognizance completion probability than a scheme using classical estimate of the number of tags separately.

• Smart Media Journal
• /
• v.10 no.3
• /
• pp.39-47
• /
• 2021

Enhanced index tracking is a problem of optimizing the objective function to generate returns above the index based on the index tracking that follows the market return. In order to avoid problems such as large transaction costs and illiquidity, we used a method of constructing a portfolio by selecting only some of the stocks included in the index. Commonly used enhanced index tracking methods tried to find the optimal portfolio with only one objective function in all tested periods, but it is almost impossible to find the ultimate strategy that always works well in the volatile financial market. In addition, it is important to improve generalization performance beyond optimizing the objective function for training data due to the nature of the financial market, where statistical characteristics change significantly over time, but existing methods have a limitation in that there is no direct discussion for this. In order to solve these problems, this paper proposes ensemble learning that composes a portfolio by combining several objective functions and a 3-stage portfolio selection algorithm that can select a portfolio by applying criteria other than the objective function to the training data. The proposed method in an experiment using the S&P500 index shows Sharpe ratio that is 27% higher than the index and the existing methods, showing that the 3-stage portfolio selection algorithm and ensemble learning are effective in selecting an enhanced index portfolio.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.18 no.6
• /
• pp.389-394
• /
• 2017

In a conventional ship structure, stiffeners with an asymmetric section, such as inverted angles, are used widely despite the disadvantage of strength compared to the stiffeners with a symmetric section, such as a built-up T. On the other hand, T-type built-up members are attracting more attention than L-type inverted angles due to the increased size of ships. The purpose of this study was to develop an optimal design program for a built-up T, and apply an evolution strategy as an optimization technique. In the optimization process, the gross thickness concept was adopted for the design variables and objective function, and the constraints are set up based on HCSR (Harmonized Common Structural Rules). Using the developed program in this study, the optimal stiffener design was carried out for 300K VLCC and 158K COT of which the orders were obtained lately. The optimal results revealed the weight reduction effect of 144 tons and 60 tons, respectively.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.16 no.6
• /
• pp.40-50
• /
• 2002

In this paper, we present a GA(Genetic Algorithm) approach to select weighting matrices of an optimal LQ(Linear Quadratic) controller for SVC(Static VAR Compensator). A SVC, one of the FACTS(Flexible AC Transmission System), constructed by a FC(Fixed Capacitor) and a TCR(Thyristor Controlled Reactor), was designed and implemented to improve the damping of a synchronous generator, as well as to control the system voltage Also, a design of LQ controller depends on choosing weighting matrices. The selection of weighting matrices which is not a trivial solution is usually carried out by trial and error. We proposed an efficient method using GA of finding weighting matrices for optimal control law. Thus, we proved the usefulness of proposed method to improve the stability of single machine-infinite bus with SVC system by eigenvalues analysis and simulation.

• Asia-pacific Journal of Multimedia Services Convergent with Art, Humanities, and Sociology
• /
• v.8 no.4
• /
• pp.251-258
• /
• 2018

In mathematics and computer science, an optimization problem is the problem of finding the best solution from feasible ones. In the context of genetic algorithm, the difficulty of an optimization problem can be explained in terms of problem epistasis. In biology, epistasis means that the phenotype of a gene is suppressed by one or more genes, but in an evolutionary algorithm it means the interaction between genes. In this paper, we experimentally show that problem epistasis and the performance of genetic algorithm are closely related. We compared problem epistasis (One-Max, Royal Road, and NK-Landscape) using a framework that quantifies problem epistasis based on Shannon's information theory, and could show that problem becomes more difficult as problem epistasis grows. In the case that a genetic algorithm finds the optimal solution, performance is compared through the number of generations, otherwise through the ratio of the fitness of the optimal solution to that of the best solution.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.18 no.5
• /
• pp.649-658
• /
• 2008

In this paper, we proposed an adaptive PSO(APSO) algorithm which changes parameter values with every recursion based on the conventional particle swam optimization(CPSO). In order to solve the optimization problem, the proposed APSO algorithm is applied to some functions, such as the De Jong function, Ackley function, Davis function and Griewank function. The superiority of the proposed APSO algorithm compared with the genetic algorithm(GA) is proved through the numerical experiment. Finally we applied the proposed algorithm to developing a function for personal magnetic field exposure based with real datas which are acquired based on the consumer research and field measuring instrument.

• Proceedings of the KIEE Conference
• /
• 2009.07a
• /
• pp.20_21
• /
• 2009

전력거래소는 국내 유일하게 한국의 전력계통을 실시간으로 감시 제어하고, 계통의 안전성을 평가 후 최적으로 운용하기 위해 2002년부터 현 EMS 시스템을 운용해 오고 있다. 초기 모든 계통해석의 근간이 되는 상태추정의 운용을 위해 계통모델과 필요 취득 데이터를 대폭 정비하였다. 현재는 154kV 무인 변전소의 탭을 추가취득 하고 모든 아날로그 취득데이터의 그룹별 가중치 조정에 의한 상태추정 등 계통해석의 정도개선 노력을 지속적으로 추진해오고 있다. 또한 상정고장 개소와 발변전소 3상 단락 고장 개소를 대폭 확대 정비하여 예상되는 모든 고장에 급전원이 신속히 대처할 수 있도록 하였다. 또한 무효전력 제어를 통한 계통손실 최소화 방안의 합리적 도출을 위한 최적화 기능의 실시간 계통운용 등 괄목한 성과를 거두었다. 본고에서는 이들 EMS 계통해석 기능을 중심으로 현재의 활용실태와 미래의 방향에 대하여 소개하고자 한다.

• Journal of IKEEE
• /
• v.22 no.2
• /
• pp.440-445
• /
• 2018

The work presented in this paper deals with a navigation problem for a multiple mobile robots in unknown dynamic environments. The environments are completely unknown to the robots; thus, proximity sensors installed on the robots' bodies must be used to detect information about the surroundings. In order to guide the robots along collision-free paths to reach their goal positions, a navigation method based on a combination of primary strategies has been developed. Most of these strategies are achieved by means of fuzzy logic controllers, and are uniformly applied in every robot. In order to improve the performance of the proposed fuzzy logic, the genetic algorithms were used to evolve the membership functions and rules set of the fuzzy controller. The simulation experiments verified that the proposed method effectively addresses the navigation problem.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2003.11a
• /
• pp.451-455
• /
• 2003

최근 LCD제품은 고유한 특장인 경박단소가 요구되면서, 기존의 Back Light Unit(BLU) 로는 대응할 수 없는 사양으로 진화되고 있다. 기존과 동일한 설계개념으로 접근시에 시장에서 요구되는 중량, 두께, 휘도의 사양을 만족시킬 수 없으며, BLU의 주요광원인 CCFL(Cold Cathod Fluorescent Lamp)의 휘도 개선 또 한 한계에 다다르고 있다. 따라서 앞으로의 BLU 의 고성능화는 최적화, 고효율화로의 개발 전개가 예상되며, LCD의 고해상도에 따른 투과율 저하를 보상하기 위한 고품질 BLU의 개발이 시급한 상황이다. 본 연구에서는 이러한 BLU의 고효율화, 고품질화를 달성하기 위한 고성능 도광판 개발과 관련하여, 실물 제작에 앞서 광학시뮬레이션을 통한 이론적 접근을 수행하였다. 연구 결과, 상측에서 정각 $90^{\circ}$ 에 높이 $50{\mu}m$ 하측에서 정각 $80^{\circ}$ 높이 $28{\mu}m$일때 평균조도가 71.52W/m^2 구현됨을 알 수 있다. 이 결과를 바탕으로 통상의 인쇄 방식 도광판에 비해서 약 20% 정도의 휘도향상이 가능함을 알 수 있었다. 또한 차후 본 결과를 바탕으로 한 실물 제작을 통해 설계 시뮬레이션 결과와의 비교를 통해서 정확한 예측이 가능한 시스템을 구현함을 목적으로 하였다.