최근에 인공생명이나 진화적 계산론이라는 이름의 새로운 지능정보처리 방식이 미국과 일본을 중심으로 활발히 연구되고 있다. 이것은 지금까지 개별적으로 제안된 두뇌의 가소성이나 개체의 발생, 적응과 진화 등 생물의 특성으로부터 파생된 모형들을 총동원하여 정보처리의 새로운 가능성을 모색하고자 하는 것이다. 본 논문에서는 인공생명의 연구가 어떻게 시작되었으며, 현재의 기술수준이 어느 정도인지에 대하여 소개하고자 한다. 아울러, 인공생명으로부터 가능한 새로운 형태의 정보처리 기능창출을 목표로 하는 연구동향을 살펴보고 앞으로의 방향을 전망해 본다.

뇌는 신경세포로 이루어진 거대한 시스템이다. 이러한 뇌의 특징은 자기조직 시스템이면서 외계의 정보구조에 맞추어서 자신의 능력을 높일 수 있다는 것이다. 또한 뇌는 병렬정보처리 방식을 대폭적으로 채용한 시스템으로서 제어기구가 전체적으로 분산되어 있다. 이러한 뇌의 동작은 구조적으로 안정적이며 그 구성소자가 어느 정도 파괴되더라도 우수한 동작특성을 유지할 수 있다. 이것은 뇌에 있어서 정보가 거시화 및 분산화 되어 있다는 증거이며, 연상기억과 내용 어드레스 기억 등과 같은 탁월한 기억방식을 실현할 뿐만 아니라 망각능력도 가지고 있다. 현실의 뇌 그 자체를 조사하는 것이 어려운 상황에서는 뇌에 관한 여러 가지 모델을 만들고 이 모델을 구체적으로 상세히 조사함으로써 현실의 뇌를 이해하고자하는 방법이 중요시 된다. 본 강연에서는 이러한 구성적 방법론의 필요성 및 뇌의 생리학적 측면, 뇌의 모델로서의 측면 그리고 신경회로망의 발전단계와 뇌 과학의 세계적 연구동향에 관하여 살펴본다.

GENIE is a learning-based engine for building intelligent systems. Learning in GENIE proceeds by incrementally modeling its human or technical environment using a neural network and a genetic algorithm. The neural network is used to represent the knowledge for solving a given task and has the ability to grow its structure. The genetic algorithm provides the neural network with training examples by actively exploring the example space of the problem. Integrated into the training examples by actively exploring the example space of the problem. Integrated into the GENIE system architecture, the genetic algorithm and the neural network build a virtually self-teaching autonomous learning system. This paper describes the structure of GENIE and its learning components. The performance is demonstrated on a robot learning problem. We also discuss the lessons learned from experiments with GENIE and point out further possibilities of effectively hybridizing genetic algorithms with neural networks and other softcomputing techniques.

This paper presents a short introduction to fuzzy measures and fuzzy integrals for providing an useful understanding of articles related on fuzzy measure theory and its applications. A brief overview of the basic concepts of systems, models, uncertainty, fuzzy measures and fuzzy integrals is provided. And terminology and notation frequently used in the discussion on the topic are introduced.

In this paper, we will define Choquet integrals of set-valued functions. Then, we show some properties of Choquet integrals of set-valued functions.

In this paper, we introduce two fuzzy convergence structures, fuzzy convergence and fuzzy limiterung, and obtain a relationship between them. We also consider relationships between fuzzy limit space and pseudotopological convergence space.

We investigate some of algebraic properties of T-generalized state machines, T-generalized transformation semigroups, coverings of T-generalized state machines and T-generalized transformation semigroups.

In this paper, we generalize the concept of fuzzy S-closed spaces due to Mukherjee and Ghosh [8] into fuzzy bitopological setting and investigate some of its properties using the concepts of (${\tau}_i$, ${\tau}_j$)-semi-closure and related notions in fuzzy setting.

We introduce fuzzy ordered filter, fuzzy weakly implicative ordered filter and fuzzy implicative ordered filter of implicative commutative semigroups and prove and some results.

In this paper, we first introduce fuzzy less strongly irresolute, fuzzy preless strongly semiopen and fuzzy pre-less strongly semiclosed mappings on fuzzy topological space, and establish their various characteristic properties. Finally, we introduce and study fuzzy less strongly semi-connectedness with the help of fuzzy less strongly semiopen sets and fuzzy less strongly semi-q-neighborhoods.

In this note we will discuss extension of fuzzy Lie subalgebra and fuzzy Lie ideals of a Lie algebra L on universal enveloping algebra U(L) of L and will study some relations among them.

고밀도 집적회로(VLSI)의 설계 과정에 있어 테스트(test)는 매우 중요한 과정으로서, 회로내의 결함(fault)을 찾기 위해 일련의 입력값을 넣어 그 출력값으로 고장 여부를 판단한다. 회로의 테스트를 위하여 사용되는 일련의 입력값을 테스트패턴(test pattern)이라 하며 최고 2n개의 테스트패턴이 생성될 수 있다. 그러므로 얼마나 작은 테스트패턴을 사용하여 회로의 결함 여부를 판단하느냐가 주된 관점이 된다. 기존의 테스트 패턴 생성 알고리즘인 휴리스틱(heuristic)조건에서 가장 큰 문제점은 빈번히 발생하는 백트랙(backtrack)과 이로 인한 시간과 기억장소의 낭비이다. 본 논문에서는 이러한 문제점을 보완하기 위해 퍼지 기법을 이용한 새로운 알고리즘을 제안한다. 제안된 기법에서는 고장신호 전파과정에서 여러개의 전파경로가 존재할 때, 가장 효율적인 경로를 선택하는 단계에서 퍼지 관계곱(Fuzzy Relational Product)을 이용한다. 이 퍼지 기법은 백트랙 수를 줄이고 기억장소와 시간의 낭비를 줄여 테스트 패턴 생성의 효율을 증가시킨다.

Hydroponics is to grow plants, not in soil but in water which the quantity of necessary chemical food can be controlled. In this paper, this is designed in the automatic system. The closed culture reduces cost of production and produces a many kinds of agricultural products in a confined place. An adaptive fuzzy control in the best method to solve and to overcome parametric uncertainties and non-linearity of the controlled system. A hydroponics automation system which is able to overcome these control problems. It is used in implementation of the hydroponics automation system. The performance is analyzed through an experiment in which the new adaptive fuzzy control method is applied to the automatic control of tomato hydroponics.

연관규칙(Association Rule)은 데이터 베이스에 존재하는 속성들 사이의 관계를 기술하는 것으로, 간단하면서도 사용자에게 많은 정보를 줄 수 있다. 그러나, 지금까지는 이진 데이터베이스에 존재하는 연관규칙의 발견에 대해서 주로 연구되어 왔으며, 실수값 속성을 갖는 데이터에 관한 연구는 미비하였다. 본 논문에서는 퍼지집합을 이용하여 실수값 사이에 존재하는 연관규칙을 기술하고, 그것을 찾아내는 방법을 제시한다. 제시하는 방법은 사용자에 의해서 정의된 언어항을 이용하여, 실수값 속성을 가진 데이터를 이진 데이터로 재구성한다. 그리고 재구성된 이진 데이터에 기존의 연관규칙 발견 방법을 이용하여 연관규칙을 찾아내고, 찾아진 연관규칙을 정의된 언어항을 이용하여 다시 기술한다.

본 논문은 Building, 아파트, 병원 호텔 등의 건물의 급수 System으로서 최근 대두되고 있는 Bosster Pump System에 관한 것으로서, 제품의 주요 특징 및 제어 알고리즘을 소개하고 특히 최종 User에게 쾌적한 급수 환경을 제공하기 위한 주 제어 기능인 일정 예측 최종 압력 제어를 PID 및 Fuzzy 제어이론을 이용하여 구현하였는데, 그 적용 알고리즘을 소개하고, 실제 제어 실험을 통해 PID제어와 Fuzzy 제어를 비교하였다.

In this paper, a fuzzy controller is proposed for the operation of stoker-type refuse incinerator with many kinds of uncertain factors. To build the exact mathematical model is very difficult because of the variation of physical/chemical properties of refuse as a fuel and the complexity of the combusiton process. The fuzzy controller consists of fuzzy sensor, fuzzy decision maker and tracking part. The rules based on the professional operators empirical knowledge are made for the control of the boiler evaporation rate, emission gas and refuse throughput. For the performance measure of the proposed fuzzy controller, the model of the incinerator is constructed and the simulation results are given.

In this paper, we suggest a framework to represent the fuzziness in knowledge base as a perspective of the object-oriented paradigm. We divide the knowledge base in two parts. One is the object-base that stores the fuzzy propositions and the explanatory databases. The other is the rule-base that manages the rules between the fuzzy propositions. As the first step, we have to develop a new fuzzy object model that gives an easy way to represent the fuzzy propositions, that is, the fuzzy knowledge in the real world.

클러스터의 타당성 정도를 계산하기 위한 측정자로서, 퍼지 분할된 데이터의 서로 다른 클래스 사이의 분리성과 한 클래스안에서의 밀접성의 비율, G를 정의하였다. 본 논문에서는 이렇게 정의된 G로부터, 각 클러스터가 가지는 데이터 수의 차이점을 고려하여 하나의 데이터 집합에 대하여 서로 다른 분할들을 비교할 수 있도록 하기 위하여, IG를 재정의하였다. 기존의 클러스터 타당성 전략은 클러스터 수의 함수로서, 주어진 척도의 값을 계산하여 기록한 후 그 값의 변화가 가장 큰 경우를 최적의 클러스터의 수로서 선택하였다. 이때 그 값의 변화를 고려하기 위한 주관적인 해석이 필요하게 된다. 본 논문에서는 주관적인 해석 없이 IG를 이용하여 최적의 클러스터 수를 결정하기 위한 방법을 제안하고자 한다. 제안된 방법은 널리 알려진 Iris data와 서로 다른 클러스터 인구수를 가지는 가상의 데이터 집합에 적용하여 그 타당성을 보인다.

In order to solve the specific problems in system, the implementing of knowledge base suitable to the system architecture has to be expressed and leads to the resoning of the new facts. Not only the simple knowledge based on the bivalent logic but also the representation and acqusition of knowledge related to vague and imprecise linguistic variable is necessary to the knowledge in the real world,i.e the knowledge of which facts and rules are composed. In this point of view, this survey implements the inference and retrieval of the fuzzy linguistic variables so that it can process the ambiguous information which appears the imprecision of the linguistic boundary.

This paper examines the realtionship between Multidimensional linear interpolation (MDI) and fuzzy reasoning, and shows that an MDI is a special form of Tsukamoto's fuzzy reasoning. From this result, we find a new possibility of defuzzification scheme.

가전제품에 사용돠고 있는 단상유도전동기의 가변속제어를 통해 다양한 소비자의 요구조건에 만족하는 제품을 개발하는 것이 중요한 문제로 대두되고 있다. 이러한 가변속제어에 필요한 속도정보를 피이드백받기 위해 유도전동기의 입력전압과 전류를 이용하여 속도추정기를 Adaptive Network Fuzzy Inference System을 이용하여 개발하였다.

Conventional fuzzy systems have serious problems in dealing with the nonlinear approximations on high-dimensional spaces due to the explosive increase of the number of fuzzy IF-THEN rules. In order to avoid such problems, this paper proposes a tree-structured fuzzy system in which semi-local basis functions form its basic elements, and develops a training algorithm for the proposed system based on the evolution program and LMS rule. The experimental studies demonstrate the effectiveness of the developed methodology.

A robust adaptive tracking control architecture is proposed for a class of continuous-time nonlinear dynamic systems for which an explicit linear parameterization of the uncertainty in the dynamics is either unknown or impossible. The architecture employs an adaptive fuzzy system to compensate for the uncertainty of the plant. In order to improve the robustness under approximation errors and disturbances, the proposed architecture includes deadzone in adaptation laws. Unlike the previously proposed schemes, the magnitude of approximate errors and disturbances is not required in the determination of the deadzone size, since it is estimated using the adaptation law. The proposed algorithm is proven to be globally stable in the Lyapunov sense, with tracking errors converging to the proposed architecture.

In this paper, we proposed an optimization method of the membership function and the numbers of fuzzy rule base for the stabilization controller of the inverted pendulum system by genetic algorithm(GAs). Conventional methods to these problems need to an expert knowledge or human experience. The proposed genetic algorithm method will tune automatically the input-output membership parameters and will optimize their rule-base.

Generally, when we control the robot, we should calculate exactly Inverse Kinematics. However, Inverse Kinematics calculation is complex and it takes much time for the manipulator to control in real-time. Therefore, the calculation of Inverse Kinematics can result in significant control delay in real time. In this paper, we will present that Inverse Kinematics can be calculated through Fuzzy Logic Mapping, Based on an exact solution through fuzzy reasoning instead of Inverse Kinematics calculation Also, the result provides sufficient precision and transient tracking error can be controlled based on a fuzzy adaptive scheme proposed in this paper. Based on the Denavit-Hartenberg parameters specification, after the Jacobian matrix of arbitrary manipulator is calculated, we will construct Fuzzy Inverse Kinematics Mapping(FIKM) using fuzzy logic and represent a good control efficiency through simulation of 2-DOF manipulator.

In the paper is proposed a hierarchical self-learning fuzzy controller for balancing and position control of an circular inverted pendulum system. To stabilize the pendulum at a specified position, the hierarchical fuzzy controller consists of a supervisory controller, a self-learning fuzzy controller, and a forced disturbance generator. Simulation example shows the effectiveness of the proposed method.

This paper describes the tip displacement of a flexible miniature arm controlled by the piezoelectric bimorph cells cemented on the surface of the arm. The arm is driven by the torques generated by the cells, and the endpoiht of the arm is controlled so that it moves in synchrony with the fluctuation of the target and maintains a constant distance to the surface of the traget. The voltage applied to the cells is controlled by a feedback signal composed of the tip displacement and the velocity. A theoretical solution is obtained by considering the cell-arm system as a stepped beam and applying time-discrete method to the governing equations of the system. The results are good agreement for a wide range of physical paramehers involved.

This work discusses simulation results for the fuzzy logic controller tested the project“Fuzzy Ramp Metering Algorithm Implementation.”The performance objectives were, in order of priority, to maximize total vehicle-miles, maximize mainline speeds, and minimize delay per vehicle while maintaining an acceptable ramp queue. In the fuzzy logic controller, the sensors from the on-ramps were helpful in maintaining reasonable ramp queue and mainline congestion because it considered these factors simultaneously. Each metered ramp had a parameter input file, which allowed the controller to be modified without recompiling the software. Consequently, maintenance costs should be minimal.

This paper presents a fuzzy structural modeling using the concept of Saaty's pairwise comparison matrix. The merit of this approach can be reduced the number of pairwise comparison.

냉장실의 온도분포의 추정은 냉장고의 과냉현상 방지, 운전효율 향상을 위해 필요할 뿐만이 아니라 최근 많은 제품에서 제공하고 있는 집중냉각기능을 실현하는데 있어서도 필수 불가결한 과정이라 할 수 있다. 본 연구에서는 냉장실내 온도분포추정에 있어서의 문제점을 개관하고, 온도분포추정을 위한 퍼지불감대(Fuzzy Dead Zone)를 갖는 퍼지동정모델(Fuzzy Identification Model)을 제안한다. 또한, 얻어진 모델을 이용하여 냉장실내 온도센서의 최적위치에 관하여 고찰한다.

Electro sensitive traffic system can't consider passenger car unit, so, it causes start up delay time and passenger waiting time. In this paper, it antecedently creates optimal traffic cycle of passenger car unit at the bottom traffic intersection. But, sometimes it can make mistakes due to changes in car weight, car speed, and control of feed-back data. Moreover, to prevent spillback, it can adapt control even though upper traffic intersection has a different saturation rate, road length, road slope and road width.

This paper propose a method for pattern recogniton using spectrum analyzer and fuzzy ARTMAP. Contour sequences obtained from 2-D planar images represent the Euclidean distance between the centroid and all boundary pixels of the shape, and are related to the overall shape of the images. The Fourier transform of contour sequence and spectrum analyzer are used as a means of feature selection and data reduction. The three dimensional spectral feature vectors are extracted by spectrum analyzer from the FFT spectrum. These Spectral feature vectors are invariant to shape translation, rotation, and scale transformations. The fuzzy ARTMAP neural network which is combined with two fuzzy ART modules is trained and tested with these feature vectors. The experiments include 4 aircrafts and 4 industrial parts recognition process are presented to illustrate the high performance of this proposed method in the ion problems of noisv shapes.

ATM망은 실제로 이용 가능한 대역폭 이상을 할당하는 통계적 다중화(statistical multiplexing) 기법을 사용하므로 망을 통한 트래픽 흐름을 적절히 관리하지 못하면 혼잡(congestion), 셀 손실, 망의 성능 저하 등을 야기하게 된다. 이러한 상황을 예방하고 셀의 도착 시간 버스트(burstiness)를 줄이며 셀 손실 특성을 개선하여 망의 성능을 증가시키기 위하여, 트래픽의 형태 제어 방법을 제안한다. 트래픽 형태 제어 파라미터 값의 역전파 신경망을 적용하여 예측되며, 이 예측된 값들에 의해 형태 제어 방법을 수행한다. 제안된 형태 제어 기법의 성능은 Poisson 트래픽 입력에 대한 컴퓨터 시뮬레이션에 의해 얻어지며, 멀티플렉서에서의 최대 버퍼 크기를 측정하여 성능을 평가하였다.

During the navigation of mobile robot, one of the essential task is to determination the absolute location of mobile robot. In this paper, we proposed a method to determine the position of the camera from a landmark through the visual image of a quadrangle typed landmark using neural network. In determining the position of the camera on the world coordinate, there is difference between real value and calculated value because of uncertainty in pixels, incorrect camera calibration and lens distortion etc. This paper describes the solution of the above problem using BPN(Back Propagation Network). The experimental results show the superiority of the proposed method in comparison to conventional method in the performance of determining camera position.

One of the consideration issue in implementation and maintanence of CIM(Computer Integrated Manufacturing) database is exchange and sharing of information between heterogeneous databases. For efficient operating of SIM systems, it must be able to organize and to manage the information. In this paper, we propose method that can make enhance the efficiency of CIM database, classfying the data in database using self organize neural network to each database systems, and computing between classfied heterogeneous database using extended operator that is defined.

The Inverted Pendulum has been one of most popular nonlinear dynamic systems for the exploration of control techniques. This paper presents a new linear optimal control techniques and nonlinear neural network learning methods. The multiayered neural networks are used to add nonlinear effects on the linear optimal regulator(LQR). The new regulator can compensate nonlinear system uncertainties that are not considered in the LQR design, and can tolerated a wider range of uncertainties than the LQR alone. The new regulator has two neural networks for modeling and control. The neural network for modeling is used to obtain a more accurate model than the given mathematical equations. The neural network for control is used to overcome deficiencies by adding corrections to the linear coefficients of the LQR and by adding nonlinear effects on the LQR. Computer simulations are performed to show the applicability and a more robust regulator than the LQR alone.

In this paper, Plasma-Enhanced Chemical Vapor Deposition (PECVD) modeling using Polynomial Neural Networks (PNN) has been introduced. The deposition of SiO2 was characterized via a 25-1 fractional factorial experiment, was used to train PNNs using predicted squared error (PSE). The optimal neural network structure and learning parameters were determined by means of a second fractional factorial experiment. The optimized networks minimized both learning and prediction error. From these PNN process models, the effect of deposition conditions on film properties has been studied. The deposition experiments were carried out in a Plasma Therm 700 series PECVD system. The models obtained will ultimately be used for several other manufacturing applications, including recipe synthesis and process control.

본 논문에서는 BP(Back Propagation)에 비해서 빠른 학습시간과 다른 경쟁학습 신경회로망 알고리즘에 비해서 비교적 우수한 성능으로 패턴인식 등에 많이 이용되고 있는 LVQ(Learning Vector Quantization) 알고리즘의 성능을 향상시키기 위한 방법을 논의하고자 한다. 일반적으로 LVQ는 음(negative)의 학습을 하기 때문에 초기 가중치가 제대로 설정되지 않으면 발산할 수 있다는 단점이 있으며, 경쟁학습 계열의 신경망이기 때문에 출력 층의 뉴런 수에 따라 성능에 큰 영향을 받는다고 알려져 있다.[1]. 지도학습 형태를 지닌 LVQ의 경우에 학습패턴이 n개의 클래스를 가지고, 각 클래스 별로 학습패턴의 수가 같은 경우에 일반적으로 전체 출력뉴런에 대해서 (출력뉴런수/n)개의 뉴런을 각 클래스의 목표(desired) 클러스터로 할당하여 학습을 수행하는데, 본 논문에서는 각 클래스에 동일한 수의 출력뉴런을 할당하지 않고, 학습데이터에서 각 클래스의 분산을 추정하여 각 클래스의 분산을 추정분산에 비례하게 목표 출력뉴런을 할당하고, 초기 가중치도 추정분산에 비례하게 각 클래스의 초기 임의 위치 입력백터를 사용하여 학습을 수행하는 방법을 제안한다. 본 논문에서 제안하는 방법은 분류하고자 하는 데이터에 대해서 필요한 최적의 출력뉴런 수를 찾는 것이 아니라 이미 결정되어 있는 출력뉴런 수에 대해서 각 클래스에 할당할 출력 뉴런 수를 데이터의 추정분산에 의해서 결정하는 것으로, 추정분산이 크면 상대적으로 많은 출력 뉴런을 할당하고 작으면 상대적으로 적은 출력뉴런을 할당하고 초기 가중치도 마찬가지 방법으로 결정하며, 이렇게 하면 정해진 출력뉴런 개수 안에서 각 클래스 별로 분류의 어려움에 따라서 출력뉴런을 할당하기 때문에 미학습 뉴런이 줄어들게 되어 성능의 향상을 기대할 수 있으며, 실험적으로 제안된 방법이 더 나은 성능을 보임을 확인했다.initially they expected a more practical program about planting than programs that teach community design. Many people are active in their own towns to create better environments and communities. The network system "Alpha Green-Net" is functional to support graduates of the course. In the future these educational programs for citizens will becomes very important. Other cities are starting to have their own progrms, but they are still very short term. "Alpha Green-Net" is in the process of growing. Many members are very keen to develop their own abilities. In the future these NPOs should become independent. To help these NPOs become independent and active the educational programs should consider and teach about how to do this more in the future.단하였는데 그 결과, 좌측 촉각엽에서 제4형의 신경연접이 퇴행성 변화를 나타내었다. 그러므로 촉각의 지각신경세포는 뇌의 같은 족 촉각엽에 뻗어와 제4형 신경연접을 형성한다고 결론되었다.$/ 값이 210 $\mu\textrm{g}$/$m\ell$로서 효과적인 저해 활성을 나타내었다 따라서, 본 연구에서 빈

In this paper, a general neural-network-based connectionist model, called Fuzzy Neural Network(FNN), is proposed for the realization of a fuzzy logic control system. The proposed FNN is a feedforward multi-layered network which integrates the basic elements and functions of a traditional fuzzy logic controller into a connectionist structure which has distributed learning abilities. Such FNN can be constructed from training examples by learning rule, and the connectionist structure can be trained to develop fuzzy logic rules and find optimal input/output membership functions. Computer simulation examples will be presented to illustrate the performance and applicability of the proposed FNN, and their associated learning algorithms.

Fuzzy logic has been successfully used for nonlinear control systems. However, when the plant is complex or expert knowledge is not available, it is difficult to construct the rule bases of fuzzy systems. In this paper, we propose a new method of how to construct automatically the rule bases using fuzzy neural network. Whereas the conventional methods need the training data representing input-output relationship, the proposed algorithm utilizes the gradient of the object function for the construction of fuzzy rules and the tuning of membership functions. Experimental results with the inverted pendulum show the superiority of the proposed method in comparison to the conventional fuzzy controller.

Rainfall is one of the major and complicated elements of hydrologic system. Accurate prediction of rainfall is very important to mitigate storm damage. The neural network is a good model to be applied for the classification problem, large combinatorial optimization and nonlinear mapping. In this dissertation, rainfall predictions by the neural network theory were presented. A multi-layer neural network was constructed. The network learned continuous-valued input and output data. The network was used to predict rainfall. The online, multivariate, short term rainfall prediction is possible by means of the developed model. A multidimensional rainfall generation model is applied to Seoul metropolitan area in order to generate the 10-minute rainfall. Application of neural network to the generated rainfall shows good prediction. Also application of neural network to 1-hour real data in Seoul metropolitan area shows slightly good predictions.

This study is about the construction of algorithm for selecting Yongshin of the Four Pillars. To emulate the method the expert uses when he select the Yongshin, we introduce the Hopfield Network. The result of the simulation classified with Yongshin is presented.

Cellular Automata are a discrete mathematical system whose evolution is governed by a deterministic rule involving local interactions. In this paper, one designed and implemented the evolution system based on GUI which can analyse how random intitial stated evolve easily.

In this paper, we present a new design method to implement autoassociative memories based on BSB neural networks. With a concrete mathematical model proposed after analyzing some new qualitative properties of autoassociative memories, we reinterpret design of autoassociative memories as a constrained optimization problem and use an evolution program as an optimal search tool to solve this. The proposed method satisfies many of the criteria used to evaluate the effectivencess of a given associative memory and has improvements with respect to correctness and performance. Comparing simulation results with other methods, we demonstrate the effectiveness of the proposed method.

In this paper a new determination scheme of linear decision function is proposed. In this scheme, the weights in linear decision function is obtained by genetic algorithm. The result considering balance between clusters as well as classification error can be obtained by properly selecting the fitness function of genetic algorithm in determination of linear decision function and this has the merit in applying this scheme to the construction of binary decision tree. The proposed scheme is applied to the artificial two dimensional data and real multi dimensional data. Experimental results show the usefulness of the proposed scheme.

The application of genetic algorithms to fuzzy rule generation holds a great deal of promise in overcoming difficult problems in fuzzy systems design. There are some aspects to be considered when genetic algorithms are used for generating fuzzy rules. In this paper, we will present an aspect about the control surface constructed by the resultant rules. In the extensive simulations, an important observation that the rules searched by genetic algorithms are randomly scattered is made and a solution to this problem is provided by including a smoothness cost in the objective function. We apply the fuzzy rules generated by genetic algorithms to the fuzzy truck backer-upper control system and compare them with the rules made by an expert.

Collision avoidance is a method to direct a mobile robot without collision when traversing the environment. This kind of navigation is to reach a destination without getting lost. In this paper, we use a genetic algorithm for the path planning and collision avoidance. Genetic algorithm searches for path in the entire, continuous free space and unifies global path planning and local path planning. It is a efficient and effective method when compared with traditional collision avoidance algorithm.

Fuzzy logic controllers have been shown better performance than conventional ones especially in highly nonlinear plants. These results are caused by the nonlinear fuzzy rules were not sufficient to cope with significant uncertainty of the plants and environment. Moreover, it is hard to make fuzzy rules consistent and complete. In this paper, we employed a predictive neural network to enhance the nonlinear inference capability. The predictive neural network generates predictive outputs of a controlled plant using the current and past outputs and current inputs. These predictive outputs are used in terms of fuzzy rules in fuzzy inferencing. From experiments, we found that the predictive term of fuzzy rules enhanced the inference capability of the controller. This predictive neural network can also help the controller cope with uncertainty of plants or environment by on-line learning.

This paper proposes a scheme of initial cluster center selection in FCM algorithm using the genetic algorithms. The FCM algorithm often fails in the search for global optimum because it is local search techniques that search for the optimum by using hill-climbing procedures. To solve this problem, we search for a hypersphere encircling each clusters whose parameters are estimated by the genetic algorithms. Then instead of a randomized initialization for fuzzy partition matrix in FCM algorithm, we initialize each cluster center by the center of a searched hypersphere. Our experimental results show that the proposed initializing scheme has higher probabilities of finding the global or near global optimal solutions than the traditional FCM algorithm.

This paper proposes the method to recognize of time series data based on the chaotic feature extraction. Features extract from time series data using the chaotic time series data analysis and the pattern recognition process is using a neural network classifier. In experiment, EEG(electroencephalograph) signals are extracted features by correlation dimension and Lyapunov experiments, and these features are classified by multilayer perceptron neural networks. Proposed chaotic feature extraction enhances recognition results from chaotic time series data.

This paper describes the structure Identification of nonlinear function using Adaptive Neuro-Fuzzy Inference Technique(ANFIS). Nonlinear mapping relationship between inputs and outputs were modeled by Sugeno-Takaki's Fuzzy Inference Method. Specially, the consequent parts were identified using a series of 1st order equations and the antecedent parts using triangular type membership function or bell type ones. According to learning Rules of ANFIS, adjustable parameters were converged rapidly and accurately.

In this paper, an optimal identification method using fuzzy-neural networks is proposed for modeling of nonlinear complex systems. The proposed fuzzy-neural modeling implements system structure and parameter identification using the intelligent schemes together with optimization theory, linguistic fuzzy implication rules, and neural networks(NNs) from input and output data of processes. Inference type for this fuzzy-neural modeling is presented as simplified inference. To obtain optimal model, the learning rates and momentum coefficients of fuzz-neural networks(FNNs) and parameters of membership function are tuned using genetic algorithm(GAs). For the purpose of its application to nonlinear processes, data for route choice of traffic problems and those for activated sludge process of sewage treatment system are used for the purpose of evaluating the performance of the proposed fuzzy-neural network modeling. The show that the proposed method can produce the intelligence model w th higher accuracy than other works achieved previously.

본 논문에서는 유전자 알고리즘을 이용하여 퍼지 제어기를 위한 최적 소속함수와 제어 규칙들을 자동으로 생성하는 방법을 제안한다. 제안한 방법은 효과적인 염색체 암호화 방법을 이용하여 소속함수의 표현 해상도가 증가하여도 소속함수의 언어항의 개수를 일정하게 유지하여 제어 규칙을 표현하는 염색체의 길이가 크게 늘어나지 않도록 한다. 또, 소속함수의 언어항의 개수가 서로 다른 염색체에 대해서도 개선된 교배 및 돌연변이 연산자를 이용하여 효과적으로 유전자 연산을 적용할 수 있게 한다. 본 논문에서는 제안된 방법을 퍼지 제어기의 자동 생성 방법의 평가 문제로 널리 이용되는 트럭 후진 주차 문제에 적용하여 성능을 평가한다.

In this paper, we propose a design method for nonlinear SISO system using Takagi-Sugeno fuzzy model and Genetic Algorithm. Our method can reduce the number of design parameters and has advantage of small search space of Genetic Algorithm. The proposed nonlinear controller, which can be implemented by fuzzy controller and simple nonlinear controller, cancels the original nonlinear dynamics and gives the optimal nonlinear dynamics. We illustrated the performance of the proposed controller by simple simulation example.

The machine-part cell formation means the grouping of similar parts and similar machines into families in order to minimize bottleneck machines, bottleneck parts, and inter-cell part movements in cellular manufacturing systems and flexible manufacturing systems. The cell formation problem is knows as a kind of NP complete problems. This paper briefly introduces the cell-formation problem and proposes a cell formation method based on the Kohonen's self-organizing feature map which is a neural network model. It also shows some experiment results using the proposed method. The proposed method can be easily applied to the cell formation problem compared to other meta-heuristic based methods. In addition, it can be used to solve large-scale cell formation problems.

Automatic gauge control algorithm in rolling process is composed of several functions. Among them feedforward control method is used to compensate irregularity of input strip thickness before rolling process. Since it's very difficult to get an explicit relation between the degree of irregularity of input strip and manipulated variables, approximate linear equation like straight line is used in real system. Furthermore parameters included in such static equation should be changed by characteristics of input strip and modified by roll states. Despite this problem, rolling process use variables in feedfroward controller as a constant. Therefore this problem increases the possibilities of irregularity of thickness control. This paper presents an algorithm which can properly infer present states of process and intelligently manipulate the parameter of feedforward controller.

In this paper, we propose a new method of the communication system for cooperative behavior and control in a collective autonomous mobile robots. A communication function among the collective robots is essential to intelligent cooperation. In general, global communication is effective for small number of robots. However when the number of robot goes on increasing, this becomes difficult to be realized because of limited communication capacity and increasing amount of information to handle. And also the problems such as communication interference and improper message transmission occur. So we propose local communication system based on infrared sensor to realize the cooperative behavior and control as the solution of above problem. It is possible to prevent overflow of information and exchange of complex information by combining communicate a specific robot. At last we verify the effectiveness of the proposed communication system from example of cooperative behavior.

Rough Set theory suggested by Pawlak has a property that it can describe the degree of relation between condition and decision attributes of data which don't have linguistic information. In this paper, by using this ability of rough set theory, we define a occupancy degree which is a measure can represent a degree of relational quantity between condition and decision attributes of data table. We also propose a method that can find an optimal fuzzy rule table and membership functions of input and output variables from data without linguistic information and examine the validity of the method by modeling data generated by fuzzy rule.

The sign language is a method of communication for deaf person. For sign communication, sign language and manual alphabet are used continuously. In this paper is proposed a system which recognize Korean sign language(KSL) and Korean manual alphabet(KMA) continuously. For recognizing KSL and KMA, basic elements for sign language, namely, the 14 hand directions, 23 hand postures, and 14 hand orientations are used. At first, this system recognize current motion state using speed and change of speed in motion by state automata. Using state, basic element classifiers using Fuzzy Min-Max Neural Network and Fuzzy Rule are executed. Meaning of signed gesture is selected by using basic elements which was recognized.

In this paper, we construct rough relational database model using approximation concepts of rough set. Also, we analyze the relation between objects, attributes and attribute values and, propose the method that can generate flexible retrieval results.

실세계에서 존재하는 대부분의 지식은 다양한 패턴들로 구성되어 있다. 본 논문에서는 사례베이스 추론(Case-Based Reasoning : CBR)에서 다중의 의미를 갖는 불확실한 지식을 쉽게 표현할 수 있는 러프 집합을 이용하여 지식의 함축의 의미를 갖는 지식을 간략화하는 방법을 제안한다. 전문가의 지식 구조를 명확화 하는데는 많은 노력이 필요하고 지식획득의 병목현상이 일어난다. 이러한 문제점을 해결하기 위해 많은 사례의 수를 러프 집합의 성질을 이용하여 사례를 동치 클래스로 분류하여 사례의 수를 감소하므로써 CBR의 기능을 향상시킨다.