電力시스템의 短期運用을 위한 火力-揚水發電시스템의 最適運轉

李鳳容,沈建輔 弘益大學校 1985 弘大論叢 Vol.17 No.2

Large-scale electric energy storage which is economically feasible is even today possible only with pumped storage hydro plant and its merit is increasing with the increased share of base units such as nuclear power and large capacity thermal units. However, in spite of its economy only a few methods or algorithms of optimal coordination between thermal and pumped storage plants could be found in literatures. This paper proposes quite a simple concept of pumping level and peak-shaving level which minimize the total operating cost. Cost functions which contain levels explicitly can easily be obtained and it is shown that the optimization is very simple. The proposed method is demonstrated through an example system of which peak load is 2,400 [MW] and has 10 generating units. Optimal solution has resulted the saving of 22,605,000 [Won] per day.

經濟給電방식에 의한 確率的 火力運轉費 시뮬레이션 모델

李鳳容,沈建輔 弘益大學校 1987 弘大論叢 Vol.19 No.2

A new probabilistic production costing simulation models have been developed based on the real economic load dispatch. Existing methodologies are based on the zero or maximum output in case of on-off two-state model, achieving only approximate economic loadings. Since the production costing is one of the most important part in the power system planning and operation, the model refinements are continuously required. The best model must be able to simulate the real world exactly and the computing efficiencies are also reasonable. In this sense, existing models can be said to be only approximate. The one using Monte-Carlo simulation may be exact, but the computing effort is generally tremendous. This paper proposes a new concept for the probabilistic production costing simulation. The key idea lies an the fact that the real economic load dispatch can only simulate the real world operation in power system. Two models, approximate and exact, have been developed. Results are compared with the available existing ones through sample systems, and the usefulness of the proposed methods is shown.

에디전트 기법을 이용한 BBS의 불건전 기사 검색 시스템의 설계 및 구현

손봉기,남기범,이건명 충북대학교 컴퓨터 정보통신 연구소 1999 컴퓨터정보통신연구 Vol.7 No.1

BBS(Bulletin Board System)의 사용이 대중화되면서 건전하고 유용한 정보뿐만 아니라 음란물 및 불법 소프트웨어의 복제 판매 등과 같은 불건전하고 불법적인 정보가 전파될 수 있다. 특히 BBS 상의 음란물은 청소년에게 정서적으로 나쁜 영향을 줄 수 있고, 불법판매 행위는 소프트웨어 관련업계에 경제적인 피해를 줄 수 있다. 현재 BBS 서비스 업체 및 기관에서는 게시판, 자료실 등에 대해 담당직원을 두어 수 작업으로 불건전 기사를 제재하고 있는데, 이 방법은 제재하는데 한계가 있고 인력과 시간의 부담이 따른다. 본 논문에서는 에이전트 기법을 이용해 BBS 내에서의 불건전 정보를 자동으로 검색하는 시스템을 설계하여 제안한다. 이 시스템은 인력과 시간의 절약은 물론 불건전 정보가 BBS를 통해 유통되는 것을 방지하는데 크게 기여할 것이라 기대된다. As the information exchanges in BBS(Bulletin Board System) become popular, unsound and illegal information like lascivious things, illegal reproduction selling of softwares as well as sound and useful information might be prevailing. Especially lascivious things in BBS exert a harmful influence upon juveniles. At present BBS service companies and the police squad manually take sanctions against unsound articles by completely charged staff monitoring BBS. This method has limitations in thoroughly monitoring BBS and requires many human resources and long retrieval time. In this paper, we design and propose a system that automatically retrieves unsound articles in BBS by using agent technology. We expect that this system can have the advantages of saving retrieval time, reducing human resources and restraining unsound information from being distributed over BBS.

다중슬랙모선 전력조류계산법과 행렬의 반복계산식을 이용한 효율적인 상정사고 해석

심건보,이봉용 弘益大學校 科學技術硏究所 1994 科學技術硏究論文集 Vol.5 No.-

Contingency analysis requires usually many simulations in order to identify weak points in the system. Often it is time conusming although its results may be exact. This paper suggests a quite simple and efficient method with the matrix iterative techniques. Bus power equations are changed so that state quantities such as voltage magnitudes and angles can be found easily after assumed contingencies. Efficiency comparison shows that the suggested method is quite superior, 7 and 192 times better efficiencies than DC load flow and AC load flow, respectively are expected in n=100 scale system. Particulary, usual one-slack load flow is avoided to get more realistic results. Multi-slack bus load flow is introduced.

자코비안 행렬의 반복계산을 이용한 뉴톤-랍슨 전력조류계산의 계산효율 개선

李鳳容,沈建輔 弘益大學校 1988 弘大論叢 Vol.20 No.2

A load flow method by matrix iterative solutions is presented. Instead of matrix inversion or bi-factorization required for the power mismatch solution of Jacobian matrix formula, the proposed method uses iterative technique in solving matrix linear equations. Usually this method is known to show and unreliable convergence. But when it is applied to load flow problem and some caution is made, the computation efficiency can greatly be improved. The method is simpler and easier for practical implementation than the factorization technique. 4 sample systems, 5, 14, 30 and 57-Bus IEEE Test System, are investigated to de-monstrate convergence characteristics. The proposed method has been shown the larger the system becomes, the greater the computation efficiency is.

명상과 바이오피드백에 대한 동서 심리학적 고찰

이봉건 서울대학교 심리과학연구소 2001 心理科學 Vol.10 No.1

This paper purports to consider from the viewpoint of psychology of consciousness established by R. Ornstein about the aspects of meditation and biofeedback. Following this viewpoint human mind consists of the intuitive mode of consciousness as well as the analytical one traditionally emphasized in the West. So the harmony between these two modes are very important for the adaptation. The disharmony between them is considered to be the cause of maladaptation. The characteristics of Eastern esoteric traditions, i.e meditation are sustained concentration, passive attitude, etc. The effects of meditation are the emergence of intuition, and these intuitive experiences enhance the adaptation of the practitioner, i.e. produce the psychotherapeutic effect. The biofeedback training is so called the western electronic meditation. The similarities between them are considered from both the eastern, esoteric viewpoint and thee western consciousness-psychology viewpoints.

뉴톤-랍슨법에 의한 전력조류계산 프로그램 개발

李鳳容,金正勳,沈建輔 弘益大學校 1981 弘大論叢 Vol.13 No.-

A power flow solution program was developed based on the Newton-Raphson iterative technique. Although this technique has been widely adopted overseas, little attention has been given to it domestically. Memory requirement and matrix inversion are the main short-comings of this technique. In this paper, one dimensional non-zero element storage scheme was developed and achieved a great reduction in matrix memory. Further, matrix inversion was efficiently performed by using Choleski's decomposition technique. The results ar demonstrated with 5 and 11 Bus Systems. 제1장 서 론 제2장 본 론 1. 모선어드미턴스 구성 앨고리즘 2. 삼각함수의 근사화 계산 3. 뉴톤-랍슨 전력 조류계산의 개요 4. 자코비안 구성 앨고리즘 5. 촐스키의 분해법 6. 자코비안의 분해 및 수정 7. 선로의 유효 및 무효전력계산 8. 전력조류계산의 플로우챠트 제3장 사례연구 제4장 결 론 A power flow solution program was developed based on the Newton-Raphson iterative technique. Although this technique has been widely adopted overseas, little attention has been given to it domestically. Memory requirement and matrix inversion are the main short-comings of this technique. In this paper, one dimensional non-zero element storage scheme was developed and achieved a great reduction in matrix memory. Further, matrix inversion was efficiently performed by using Choleski's decomposition technique. The results ar demonstrated with 5 and 11 Bus Systems.

균형된 결정트리의 형성을 위한 군집화 알고리즘

李根壽,金奉權 안성산업대학교 1991 論文集 Vol.23 No.-

알려지지 않은 방대한 량의 자료들을 효과적으로 분류하고 검색하기 위한 시스템에서는 균형된 결정트리를 형성하는 것이 효율적인 검색을 위해 매우 중요한 관건이 된다. 그러나 기존의 군집화 알고리즘(clustering algorithm)들은 단지 그 구성원이 서로 동질이고 다른 집단과 잘 구분될 수 있는 구조를 찾아내기 위한 목적을 가지고 수행한다. 본 논문에서는 균형된 결정트리를 형성하기 위해 가능한한 형성된 집단들 간에 동질성(homogeneity)을 잃지 않으면서 등분포(equi-distributed)의 성질을 가지도록 하는 군집화 알고리즘에 대해 기술하며, 이를 위해 등분포율(EDR:Equi-Distribution Rate)을 정의한다. 이러한 알고리즘을 이용하여 형성되는 결정트리는 많은 양의 자료들을 검색하여야 하는 영상데이타베이스에서 유사성 검색(similarity retrieval)과 같은 목적을 위해 효율적으로 사용될 수 있다. The equi-distributed decision tree is very important in the system to store and retrieve a very large set of data. But the object of traditional clustering algorithms is to find the homogenious cluster structure. This paper describes an algorithm that used when designing balanced decision tree. The EDR(Equi-Distribution Rate) defined to form equi-distributed cluster. It will be use for similarity retrieval in the image database system.

퍼지 영역 지식을 이용한 데이터베이스 요약

손봉기,이우항,이건명 충북대학교 컴퓨터 정보통신 연구소 2000 컴퓨터정보통신연구 Vol.8 No.1

이 논문에서는 속성값의 일반화에 퍼지 영역 지식을 사용하는 새로운 요약 방법을 제안한다. 제안한 방법은 속성-기반 귀납 접근 방법을 따르고, 비수치 속성의 일반화에는 퍼지 개념 계층을 사용하고 수치 속성의 일반화에는 퍼지 언어항을 사용한다. 퍼지 개념 계층은 개념들 사이의 일반화 정도의 차이를 반영하며, 퍼지 언어항은 수치 속성값에 대한 언어적 요약에 사용된다. 퍼지 개념 계층과 퍼지 언어항을 사용함으로서 데이터베이스에 대한 보다 의미 있는 요약을 할 수 있다. 제안된 데이터베이스 요약방법의 응용성을 보이기 위해, 수치 및 비수치 데이터 뿐 아니라 처지 속성을 갖는 데이터에 대한 실험 과정과 결과를 보인다. 실험을 통해 제안된 방법이 퍼지 영역 지식을 이용해 데이터베이스로부터 보다 의미있는 요약을 추출한다는 것을 알 수 있다. 제안된 방법은 일반 데이터베이스뿐만 아니라 퍼지 수치 데이터를 가지는 데이터베이스 요약에도 적용될 수 있다. This paper proposes a new summarization method that uses fuzzy domain knowledge for generalization of attribute values. The method follows the attribute-oriented induction approach, and uses fuzzy concept hierarchies for nominal attribute generalization and fuzzy linguistic terms for numeric attribute generalization. Fuzzy concept hierarchies are used to reflect the difference in generalization degrees among pairs of concepts. In addition, fuzzy linguistic terms are employed to get linguistic summary for numeric attribute values. The use of fuzzy concept hierarchies and fuzzy linguistic terms helps to produce comprehensive summary for databases. In addition, it presents some experiment results to show the applicability of the proposed summarization methods. From the experiments, we observe that the proposed method can extract meaningful summaries from databases having fuzzy domain knowledge. The proposed method can be applied to both conventional databases without fuzzy quantities and databases with fuzzy numeric quantities.

공급지장비 평가에 의한 전원계획에서의 적정예비력 연구

李鳳容,沈建輔 弘益大學校 1990 弘大論叢 Vol.22 No.2

Traditionally the reliability index for generation planning has been used 0.7[days/year] in LOLP in this country. However, it seems to be time that the index should be reexamined as the scale of the power system has been grown remarkably during past decades. Reliability expression by LOLP may not be enough since it lacks information of such as severity of supply shortage. It can be complemented by other indexes, but these are not widely used due to their complication. This paper presents a result how supply shortage costs influence reserve levels of generation planning and that the appropriate level has been tried to be determined from those influences. New reliability index is suggested.