• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2000.05a
• /
• pp.220-223
• /
• 2000

타입-2 퍼지집합을 이용하여 퍼지논리시스템(Fuzzy Logic System : FLS)을 구현하기 위한 연구들이 R. I John, N. Karnik, J. Mendel 등에 의해 현재 진행되고 있다. 타입-2 집합을 이용한 타입-2 FLS은 기존의 타입-1 FLS보다 제어규칙이나 소속함순가 가지고 있는 불확실성을 표현하는데 있어서 더 효과적이다. 그러나, 타입-2 FLS 역시 타입-1 FLS이 가지고 있는 문제점인 설계시 전문가에게 의존하여 시간과 비용이 많이 소요되고, 제어기의 구성요소들을 효율적으로 생성하기가 어렵다는 문제점을 더욱 심각하게 가지고 있다. 또한, 그 문제점을 해결하기 위한 연구들도 아직 미진한 상태이다. 본 논문에서는 타입-2 FLS의 설계를 위해 유전자 알고리즘을 사용하는 방법을 제안한다. 타입-2 FLS를 설계하기 위해서는 소속함수와 제어규칙을 생성하여야 한다. 본 논문에서는 유전자 알고리즘을 사용하여 타입-2 퍼지제어규칙과 소속함수를 설계하는 방법을 제안한다. 먼저, 유전자 알고리즘에서 사용할 수 있는 유전자의 형태로 타입-2 퍼지제어규칙과 소속함수를 표현하기 위한 인코딩방법을 제안하고, 각각의 염색체를 진화시키기 위한 교차 연산자와 돌연변이 연산자를 정의한다. 그리고, 제안된 방법을 함수근사문제에 적용하여 유효성과 성능을 평가, 검증한다.

• Annual Conference of KIPS
• /
• 2002.11b
• /
• pp.1237-1240
• /
• 2002

본 논문에서는 ATM과 같은 연결 지향적 고속네트워크에서, 가중치를 가진 Flow들의 대역폭 할당과 라우팅문제에 있어 공정성과 처리량에 대하여 고려해 보았다. 가중치클 고려치 않은 Flow들에 대한 최적경로설정문제에 대하여, 기존의 QoS 서비스와 Best-Effort 서비스에서 연구된 라우팅알고리즘에서 벗어나, 본 논문은 가중치를 가진 Flow들에 대하여 Soft-QoS서비스를 지원함에 있어서 공정성과 최대 처리량을 정의하고, 또한 이를 바탕으로 가중치 최소극대 대역폭 할당과 가중치 최소극대 공정라우팅 알고리즘을 제안한다. 종단간 최적경로를 설정하는데, 최소비용으로 Bottleneck-Link를 구하고 대역폭을 할당하기 위하여 그래프 상의 노드에 두 가지 색을 사용하는 그래프문제(Graph Coloring)와 최악의 경우를 감안하면서 경로를 선택하는 최소극대화 문제(Maxmin)를 결부시켜 살펴본다. 나아가 Soft-QoS 서비스의 최대값과 최소값을 고려한 가중치를 가진 Weighted-Flow들의 대역폭 할당과 경로설정에 있어, 동적인 네트워크 환경에 보다 효율적으로 접근 가능한 근사 알고리즘을 제안한다.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2002.12a
• /
• pp.75-79
• /
• 2002

본 논문에서는 선박 디젤엔진의 속도를 조절함에 있어서 환경 변화에서도 지속적으로 만족스러운 성능을 유지하도록 시스템 파라미터 추정자, 제어기 계수 수정자를 결합한 자기동조 PID 제어기를 설계한다. 유전알고리즘을 기반으로 한 온라인 추정자가 시스템의 파라미터를 추정하면, 제어기 계수 수정자는 제안한 동조규칙으로 계수를 조정하게 된다. 이를 위해 시스템을 시간지연 1차 모델로 근사화하여 유전알고리즘 기반으로 그 파라미터를 온라인 추정하는 문제를 다룬다. 제안한 방법의 성능은 B&W사의 4L80MC 디젤엔진을 제어대상으로 퍼지모델을 얻고 모의실험을 통하여 확인한다.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.47 no.11
• /
• pp.1-6
• /
• 2010

In Sensor Networks, the problem of finding the optimal routing path dynamically, which passes through all terminals or nodes once per each, may come up. Providing a generalized scheme of approximations that can be applied to the kind of problems, and formulating the bounds of the run time and the results of the algorithm made from the scheme, one may evaluate mathematically the routing path formed in a given network. This paper, dealing with Euclidean TSP(Euclidean Travelling Sales Person) that represents such problems, provides the scheme for constructing the approximated Euclidean TSP by parallel computing, and the ground for determining the difference between the approximated Euclidean TSP produced from the scheme and the optimal Euclidean TSP.

• Journal of KIISE:Computer Systems and Theory
• /
• v.33 no.10
• /
• pp.768-776
• /
• 2006

Given an unstructured point set, we use an MLS (melting least-squares) approximation to estimate the local curvatures and their derivatives at a point by means of an approximation surface Then, we compute neighbor information using a Delaunay tessellation. feature points can then be detected as zero-crossings, and connected using curvature directions. Also this approach has a fast computation time than previous methods, which based on triangle meshes. We demonstrate our method on several large point-sampled models, rendered by point-splatting, on which the feature lines are rendered with line width determined from curvatures.

• Computational Structural Engineering
• /
• v.9 no.1
• /
• pp.141-149
• /
• 1996

Efficient approximate reanalysis techniques based on substructuring are presented. In most optimal design problems, the analysis precedure must be repeated many times. In particular, one of the main obstacles in the structural optimization systems is high computational cost and time required for the repeated analysis of large-scale structural systems. The purpose of this paper is to show how to evaluate efficiently the sturctural behavior of new designs using information from the previous ones, instead of the multiple repeated analysis of basic equations for successive modification in the optimal design. The proposed reanalysis method is a combined Taylor series expansion and reduced basis method based on substructuring. Several numerical examples illustrate the effectiveness of the method.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.18 no.2
• /
• pp.123-131
• /
• 2005

According to ow previous study, we confirmed That the Petrov-Galerkin natural element method(PG-NEM) completely resolves the numerical integration inaccuracy in the conventional Bubnov-Galerkin natural element method(BG-NEM). This paper is an extension of PG-NEM to two-dimensional geometrically nonlinear problem. For the analysis, a linearized total Lagrangian formulation is approximated with the PS-NEM. At every load step, the grid points ate updated and the shape functions are reproduced from the relocated nodal distribution. This process enables the PG-NEM to provide more accurate and robust approximations. The representative numerical experiments performed by the test Fortran program, and the numerical results confirmed that the PG-NEM effectively and accurately approximates The large deformation problem.

• Journal of Sensor Science and Technology
• /
• v.9 no.6
• /
• pp.440-447
• /
• 2000

This paper describes a practical system to classify material temperature responses by composition of curve fitting and neuro-fuzzy inference. There are problems with a classification system which utilizes temperature responses. It requires too much time to approach the steady state of temperature response and it has to be filtered to remove the noise which occurs in experiments. Thus, this paper proposes a practical method using curve fitting only for transient state to remove the above problems of time and noise. Using the neuro-fuzzy system, the thermal conductivity of the material can be inferred on various ambient temperatures. So the material can be classified via its inferred thermal conductivity. To realize the system, we designed a contact sensor which has a similar structure with human finger, implemented a hardware system, and developed a classification software of curve fitting and neuro-fuzzy algorithm.

• The Journal of the Acoustical Society of Korea
• /
• v.27 no.3
• /
• pp.149-153
• /
• 2008

General underwater target detection methods adopt short-time FFT for estimate target doppler. This paper proposes the efficient target detection method, instead of conventional FFT, using approximate FFT for distributed sensor network target detection, which requires lighter computations. In the proposed method, we decrease computational rate of FFT by the quantization of received signal. For validation of the proposed method, experiment result which is applied to FFT based active sonar detector and real oceanic data is presented.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.42 no.1
• /
• pp.59-65
• /
• 2005

This paper proposes a novel TSK fuzzy modeling algorithm. Various approaches to fuzzy modeling when noise or outliers exist in the data have been presented but they are approaches to degrade effects of outliers or large noise by using loss function in the cost function mainly. The proposed algorithm is the modified version of noise clustering algorithm, and it adopts the method that does not use loss function, but method to cluster noise in a class. Noise clustering is a prototype-based clustering algorithm and it has no capability to regress. It conducts clustering of data first, and then conducts fuzzy regression. There are many algorithms to obtain parameters of premise and consequent part simultaneously, but they need to adapt the parameters obtained for more accurate approximation. In this paper, fuzzy regression is conducted with clustering by modifying noise clustering algorithm. We propose the algorithm that parameters of the premise part and the consequent part are obtained simultaneously, and the parameters obtained are not needed to adapt. We verify the proposed algorithm through simple examples and evaluate the test results compared with existing algorithms. The proposed algorithm shows robust performance against noise and it is easy to implement.