• Journal of the Korean Institute of Intelligent Systems
• /
• v.5 no.3
• /
• pp.52-63
• /
• 1995

A neuro-fuzzy algorithm is presented for nuclear reactor power control in a pressurized water reactor. Automatic reacotr power control is complicated by the use of control rods because of highly nonlinear dynamics in the axial power shape. Thus, manual shaped controls are usually employed even for the limited capability during the power maneuvers. In an attempt to achieve automatic shape control, a neuro-fuzzy approach is considered because fuzzy algorithms are good at various aspects of operator's knowledge representation while neural networks are efficinet structures capable of learning from experience and adaptation to a changing nuclear core state. In the proposed neuro-fuzzy control scheme, the rule base is formulated based ona multi-input multi-output system and the dynamic back-propagation is used for learning. The neuro-fuzzy powere control algorithm has been tested using simulation fesponses of a Korean standard pressurized water reactor. The results illustrate that the proposed control algorithm would be a parctical strategy for automatic nuclear reactor power control.

• Journal of Mechanical Science and Technology
• /
• v.15 no.12
• /
• pp.1664-1675
• /
• 2001

In this paper, the development of a new group controller for high-speed elevators is described utilizing the approach of adaptive dual fuzzy logic. Some goals of the control are to minimize the waiting time, mean-waiting time and long-waiting time in a building. When a new hall call is generated, all adaptive dual fuzzy controller evaluates the traffic patterns and changes the membership function of a fuzzy rule base appropriately. A control algorithm is essential to control the cooperation of multiple elevators in a group and the most critical control function in the group controller is an effective and proper hall call assignment of the elevators. The group elevator system utilizing adaptive dual fuzzy control clearly performs more effectively than previous group controllers.

• Journal of IKEEE
• /
• v.13 no.3
• /
• pp.32-38
• /
• 2009

Ultrasonic motor (USM) is a new type motor which is driven by the ultrasonic vibration of piezoelectric elements. It possess many useful features that electromagnetic motors do not have, such as low speed and high torque. However, ultrasonic motor has heavy nonlinear speed characteristics and is sensitive to the change of drive conditions. In order to solve these problems, we present a smooth position control scheme for ultrasonic motor using fuzzy logic control with human expertise and without the need of any precise mathematical model. A "smooth operation" consideration is included when constructing the fuzzy rule base in order to achieve a most smooth response. An experimental position control system is constructed to show the effectiveness of the proposed control scheme.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.5 no.2
• /
• pp.1-11
• /
• 2001

내진성능평가 시스템은 구조시스템의 합리적인 분류, 적절한 평가 기준, 그리고 종합적인 평가방법을 포함하여야한다. 외국의 현행 내진성능 평가방법은 데이터의 수집과 주요 평가 항목을 위한 약산식 그리고 평가 점수를 이용하여 전문가의 판단에 근거한 평가 방법을 제시하고 있다. 본 연구는 국내 건축구조물에 예비 내진평가 방법에 중점을 두고 퍼지추론 시스템에 근거한 내진평가방법의 전형을 개발한다. 평가항목의 위계는 건무의 수직, 수평방향을 불규칙성, 비대칭성, 여용성, 그리고 건물 연한을 포함한 전체적인 특성과 부재 단계에서의 상세한 평가 항목으로 구성한다. 퍼지추론방법에 대한 기존의 연구결과를 근허가혀 이용한 내진성능 평가방법에 적절히 적용하기 위하여 4가지 주요 모듈을 설정한다. (1) 퍼지 입력 (2) 퍼지에 근거한 규칙기반 (3) 퍼지추론, 그리고 (4) 퍼지출력으로 구성된다. 더욱이 개별적인 성능 수준에 종합적인 평가지수를 끌어내기 위하여 퍼지추론방법을 적용하였다.

• Advances in concrete construction
• /
• v.9 no.4
• /
• pp.407-412
• /
• 2020

Due to the influence of nonlinearity and time-variation, it is difficult to establish an accurate model of concrete frame structures that adopt active controllers. Fuzzy theory is a relatively appropriate method but susceptible to human subjective experience to decrease the performance. To guarantee the stability of multi-time delays complex system with multi-interconnections, a delay-dependent criterion of evolved design is proposed in this paper. Based on this criterion, the sector nonlinearity which converts the nonlinear model to multiple rule base of the linear model and a new sufficient condition to guarantee the asymptotic stability via Lyapunov function is implemented in terms of linear matrix inequalities (LMI). A numerical simulation for a three-layer reinforced concrete frame structure subjected to earthquakes is demonstrated that the proposed criterion is feasible for practical applications.

• Proceedings of the KIEE Conference
• /
• 2002.07d
• /
• pp.2120-2122
• /
• 2002

There are several types of control systems that use fuzzy logic controller as a essential system component. The majority of research work on fuzzy PID controller focuses on the conventional two-input PI or PD type controller. However, fuzzy PID controller design is a complex task due to the involvement of a large number of parameters in defining the fuzzy rule base. In this paper we combined conventional PI type and PD type fuzzy controller and set the initial parameters of this controller from the conventional PID controller gains obtained by Ziegler-Nichols tuning or other coarse tuning methods. After that, by replacing some of these parameters with sing1e neurons and making them to be adjusted by back-propagation learning algorithm we designed a neuro-fuzzy controller which showed good performance characteristics in both computer simulation and actual application.

• Korean Journal of Computational Design and Engineering
• /
• v.3 no.1
• /
• pp.31-39
• /
• 1998

This paper deals with a development of the system that implements a collaborative design environment with some intelligent agents on the m. Intelligent agents can design collaboratively trough an interchange of messages in their special domains. Such a collaborative design of agents is achieved on the WWW. In this paper, we propose special agents named intercessor and DCM (Dynamic Connection Manager) and suggest new connection architecture using these agents in the WWW in order to improve the pre-existed agent connection architecture. The proposed agents are developed using Java language and JATLite API. We apply the these agents to the new architecture and show some possibilities that the agent connection architecture can be extended in the WWW Agents interchange messages with others using KOML (Knowledge Query and Manipulation Language), agent communication protocol and language, and deal with message autonomously according to their rule base. Agents register and connect dynamically trough the intercessor agent, and infer from their.

• Proceedings of the KIEE Conference
• /
• 1995.07b
• /
• pp.682-684
• /
• 1995

The most effective way to improve the performance of a fuzzy controller may be to optimize look-up values. Look-up values are derived from processes used input-output scale factors, membership functions, rule base, fuzzy inference method and defuzzification. It is powerful way to modify or organize look-up table values. In this paper, We propose the look-up values self-organizing fuzzy controller(LSOFC). We use the plus-minus tuning method(PMTM), scanning values through the processes of addition and subtraction. We show the efficiency of this LSOFC by the results of simulation for nonlinear time-varying plant with unmodelled dynamics.

• ETRI Journal
• /
• v.26 no.5
• /
• pp.437-442
• /
• 2004

In this paper we address the problem of capacity optimization in a Universal Mobile Telecommunication System (UMTS) radio network. We present an optimization algorithm for finding the best settings of the antenna tilt and common pilot channel power of the base stations. This algorithm is a parametric method, based on a set of rules. We evaluated our optimization technique on a virtual network scenario with 75 cells. For this scenario we show an increase in capacity compared to the initial settings of about 60 percent.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.38 no.1
• /
• pp.74-82
• /
• 2015

Two key challenges in statistical relational learning are uncertainty and complexity. Standard frameworks for handling uncertainty are probability and first-order logic respectively. A Markov logic network (MLN) is a first-order knowledge base with weights attached to each formula and is suitable for classification of dataset which have variables correlated with each other. But we need domain knowledge to construct first-order logics and a computational complexity problem arises when calculating weights of first-order logics. To overcome these problems we suggest a method to generate first-order logics and learn weights using association analysis in this study.