• 대한전기학회논문지:시스템및제어부문D
• /
• 제49권11호
• /
• pp.632-636
• /
• 2000

Fuzzy controllers show good performance in case of the systems being nonlinear and difficult to solve. But these fuzzy controllers have problems which have to decide suitable rules and membership functions. In general, we decide those using the heuristic methods or the experience of experts. Recently, G.A. have been studied in this field. The number of rules increase exponentially when the number of input and output increase. It also makes hard to decide the rules and membership functions even though we use G.A. In this paper, we suggest parallel fuzzy controllers, and also the method to decrease the number of rules. The excellent performance of these methods is confirmed through simulations.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1998년도 하계학술대회 논문집 B
• /
• pp.566-568
• /
• 1998

In this paper, we suggest a optimal design method of Fuzzy-Neural Networks model for complex and nonlinear systems. FNNs have the stucture of fusion of both fuzzy inference with linguistic variables and Neural Networks. The network structure uses the simpified inference as fuzzy inference system and the BP algorithm as learning procedure. And we use a clustering algorithm to find initial parameters of membership function. The parameters such as membership functions, learning rates and momentum coefficients are easily adjusted using the genetic algorithms. Also, the performance index with weighted value is introduced to achieve a meaningful balance between approximation and generalization abilities of the model. To evaluate the performance index, we use the time series data for gas furnace and the sewage treatment process.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2000년도 하계학술대회 논문집 D
• /
• pp.2380-2382
• /
• 2000

Fuzzy system which are based on membership functions and rules, can control nonlinear, uncertian, complex system well. However, Fuzzy controller has problems: It is difficult to design a stable for amateur. To update the then-part membership functions of the fuzzy controller can be designed using the error back-propagation algorithm to be minimized error. Then we could be optimized the system choosing a good performance index. The proposed fuzzy controller based on neural network is applied to control an inverted pendulum for demonstration of the robustness of proposed methodology.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1995년도 하계학술대회 논문집 B
• /
• pp.652-654
• /
• 1995

This paper proposes a design of self-tuning fuzzy controller based on evolutionary computation. Optimal membership functions are round by using evolutionary computation. Genetic algorithms and evolution strategy are used for tuning of fuzzy membership function. A arbitrarily speed trajectories is selected to show the performance of the proposed methods. Simulation results show the good performance in the DC motor control system with the self-tuning fuzzy controller based on evolutionary computation.

• 대한전기학회논문지
• /
• 제41권8호
• /
• pp.938-946
• /
• 1992

A novel 2-dimensional matched filter of the parallel-jaw type using fuzzy logic is proposed for bin picking. Specifically, the averaged pixel intensity of the windowed region for the filtering is considered to be fuzzy. Also membership functions for darkness and brightness are designed by employing the intensity histogram of the image. Then a rule is given to know how much a windowed region can be a possible holdsite. Furthermore eight rules are made to determine the part orientation, where Mamdani's reasoning method is applied. The proposed technique shows better performances than that of the conventional matched filtering technique in the following senses` 1) most of holdsites determined by the proposed technique are not concentrated at the locations nearly the end of part and 2) our filter is rather insensitive to noises than the conventional method. To show the validities of our proposed technique, some experimental results are illustrated and compared with the results by conventional matched filter technique.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2000년도 추계학술대회 학술발표 논문집
• /
• pp.259-262
• /
• 2000

In this paper, an Adaptive neuro-fuzzy Inference system(ANFIS) using fuzzy min-max network(FMMN) is proposed. Fuzzy min-max network classifier that utilizes fuzzy sets as pattern classes is described. Each fuzzy set is an aggregation of fuzzy set hyperboxes. Here, the proposed method transforms the hyperboxes into gaussian menbership functions, where the transformed membership functions are inserted for generating fuzzy rules of ANFIS. Finally, we applied the proposed method to the classification problem of iris data and obtained a better performance than previous works.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 1995년도 추계학술대회 학술발표 논문집
• /
• pp.273-278
• /
• 1995

In this paper, we propose "a graph structured fuzzy system" which is able to represent the fuzzy system with a graph and optimizes the fuzzy membership functions and fuzzy rule bases using genetic algorithms. It performs the structure identification phase and parameter tuning phase simultancously through the evolutionary process. Additionally, it alleviates some of the drawbacks associated with the current fuzzy construction method with respect to the explosive increase of fuzzy rules which is inevitably encountered whenever the fuzzy systems are applied to problems with the high-dimensional input space.

• Journal of Biosystems Engineering
• /
• 제21권3호
• /
• pp.283-292
• /
• 1996

Automatic guidance of farm tractors would improve productivity by reducing operator fatigue and increasing machine performance. To control tractors within $\pm$5cm of the desired path, fuzzy and adaptive steering controllers were developed to evaluate their characteristics and performance. Two input variables were position and yaw errors, and a steering command was fed to tractor model as controller output. Trapezoidal membership functions were used in the fuzzy controller, and a minimum-variance adaptive controller was implemented into the 2-DOF discrete-time input-output model. For unit-step and composite paths, a dynamic tractor simulator was used to test the controllers developed. The results showed that both controllers could control the tractor within $\pm$5cm error from the defined path and the position error of tractor by fuzzy controller was the bigger of the two. Through simulations, the output of self-tuning adaptive controller was relatively smooth, but the fuzzy controller was very sensitive by the change of gain and the shape of membership functions. Contrarily, modeling procedure of the fuzzy controller was simple, but the adaptive controller had very complex procedure of design and showed that control performance was affected greatly by the order of its model.

• 한국정보통신학회논문지
• /
• 제16권2호
• /
• pp.228-236
• /
• 2012

본 논문에서는 빠르고 조작이 간편한 이동 수단인 이륜 역진자 로봇을 기존의 방법보다 더욱 안정적으로 밸런싱하기 위한 제어기를 설계하였다. 먼저 이륜 역진자 로봇의 제어기를 퍼지제어 구조로 선택하고, 지정된 3명의 사용자 무게에 따라 적절한 소속함수 요소 값들을 시행착오적으로 구하였다. 임의의 무게에 대한 적절한 퍼지 소속함수 요소 값을 구하기 위해 앞의 3명의 무게에 따른 퍼지 소속함수 요소 값들을 신경회로망으로 튜닝한 뒤 퍼지 제어기에 적용하여 보다 안정적인 제어가 가능하도록 제어기를 설계하였다. 설계된 제어기를 시뮬레이션 하여본 결과, 기존의 퍼지 제어기에 비해서 본 논문에서 제안한 신경회로망으로 튜닝한 퍼지제어기가 보다 안정적인 제어가 가능함을 확인할 수 있었다.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2007년도 심포지엄 논문집 정보 및 제어부문
• /
• pp.204-206
• /
• 2007

In this study, we introduce the optimization method of fuzzy inference systems that is based on Hierarchical Fair Competition-based Parallel Genetic Algorithms (HFCGA) and information data granulation, The granulation is realized with the aid of the Hard C-means clustering and HFCGA is a kind of multi-populations of Parallel Genetic Algorithms (PGA), and it is used for structure optimization and parameter identification of fuzzy model. It concerns the fuzzy model-related parameters such as the number of input variables to be used, a collection of specific subset of input variables, the number of membership functions, the order of polynomial, and the apexes of the membership function. In the optimization process, two general optimization mechanisms are explored. The structural optimization is realized via HFCGA and HCM method whereas in case of the parametric optimization we proceed with a standard least square method as well as HFCGA method as well. A comparative analysis demonstrates that the proposed algorithm is superior to the conventional methods. Particularly, in parameter identification, we use the UNDX operator which uses multiple parents and generate offsprings around the geographic center off mass of these parents.