PWM 정류초퍼 히스테리시스 전류제어의 SMPC

卓漢浩,金昌圭,金炫悳 진주산업대학교 1991 論文集 Vol.29 No.-

위치제어기는 관성이 클 경우 가속과 감속동안 동력학 에너지 변환이 이루어 져야한다. 이 경우 복합된 전류제어 PWM 정류기와 PWM 초퍼 링크로서 시스템에 전력을 변환시킬 수 있다. 따라서 본 논문에서는 다양한 위치제어 방법중 PWM정류초퍼 히스테리시스 전류제어 및 슬라이딩 기법을 이용하여 위치제어를 행하고자 한다. 이 경우에 있어서 위치제어는 다음과 같은 특성을 얻을 수 있다. 1. 조정된 정현과 전류파형을 히스테리시스 제어로 근접시킬 수 있고, 2. 슬라이딩 모드제어로써 외란, 파라메타 변화에 강인함을 얻을 수 있었다. Position controllers must handle kinetic energy exchanges during acceleratin and deceleation of large inertias. Temporary buffer storage in capacitors is expensive. The system that can manage the power exchange with ease is the integrated controlled-current PWM(pulse width modulated ) rectifier/controlled-current PWM chopper link. Bidirectional power is accomplished by bidirectional current flow in the dc link for both modulators. The three-phase current in the rectifier side are near sinusoidal in waveform at unity power factor. The performance of the system as a position controller in an enviroment of the sliding mode strategy in studied. The sliding mode strategy offers robust characteristics to the controller.

DC 전동기 제어를 위한 뉴로-퍼지 적응제어기의 설계

탁한호,김명규 진주산업대학교 1999 論文集 Vol.38 No.-

This paper aims at finding out an efficient method to control a DC Motor using Neuro-Fuzzy Adaptive Controller(NFAC). Generally parameters or operating conditions of DC Motor controllers are uncertain or unknown compared with the case of PID controller. PID controller, however, does not guarantee a good performance, which is possible only with precisely known system parameters and operating conditions. It is difficult to control an unstable nonlinear control system. Fuzzy theory has been effectively applied to this system. But it is difficult to exactly design the fuzzy controller, because membership functions and rules of a fuzzy controller are not always available. The proposed adaptive structure of the fuzzy reasoning system can be controlled more adaptively and effectively in nonlinear environment by changing input and output membership functions. The proposed fuzzy inference algorithm is named NFAC. The NFAC performance is more effective than that of the PID in the convergence time. Compared with PID, it provides an improved performance.

신경회로망을 이용한 효과적인 비선형시스템의 제어

탁한호(Han Ho Tack),이상배(Sang Bae Lee) 한국지능시스템학회 1994 한국지능시스템학회 학술발표 논문집 Vol.- No.-

유연한 로보트 매니퓰레이터의 위치제어에 관한 연구

卓漢浩,李寅龍 진주산업대학교 1996 論文集 Vol.35 No.-

In this paper, applications of the multilayer neural networks to the control of flexible robot manipulators are considered. The multilayer neural network can be used to aproximate any continuous function to any desired degree of accuracy and the weights are updaed by Gradient Method. When the flexible manipulator is rotated by motor through the fixed end, the transverse vibration may occur. The motor torque should be controlled in such a way that the motor rotates by a specified angle, while simultaneously stabilizing vibration of the flexible manipulators so that it is arrested as soon as possible at the end of rotation. Accurate control of lightweight manipulators during the large changes in configuration common to robotic tasks requires dynamic models that describe both the rigidbody motions, and the flexural vibrations. Therefore, this paper derives a linear dynamic state-space model for a single link flexible robot manipulator and composed LQR controller and inverse dynamical neuro-controller. The effectiveness of the proposed control system was proved by computer simulation.

유연성 로봇 링크의 위치제어를 위한 신경망 제어기의 설계

탁한호(Han Ho Tack),이주원(Ju Won Lee),이상배(Sang Bae Lee) 한국지능시스템학회 1997 한국지능시스템학회 학술발표 논문집 Vol.7 No.2

신경회로망을 이용한 유연한 로보트 빔의 위치제어에 관한 연구

탁한호,이상배 한국항해항만학회 1997 한국항해학회지 Vol.21 No.1

In this paper, applications of multilayer neural networks to control of flexible robot beam are considered. The multilayer nerual networks can be used to approximate any continuous function to a desired degree of accuracy and the weights are updated by Gradient Method. When a flexible beam is rotated by a motor through the fixed end, transverse vibration may occur. The motor torque should be controlled insuch a way that the motor rotates by a specified angle, while simultaneously stabilizing vibration of the flexible manipulators so that is arrested as soon as possbile at the end of rotation. Accurate control of lightweight beam during the large changes in configuration common to robotic tasks requires dynamic models that describe both rigid body motions, as well as the flexural vibrations. Therefore, a linear dynamic state-space model of for a single link flexible robot beam is derived and PD controller, LQP controller, and inverse dynamical neural networks controller are composed. The effectiveness the proposed control system is confirmed by computer simulation.

무게중심법을 이용한 무인 운반차(AGV)의 운전제어

탁한호(Han Ho Tack),권성갑(Sung Gab Kwon) 한국산업정보학회 2014 한국산업정보학회논문지 Vol.19 No.2

제조공장 또는 물류 창고의 화물운송 자동화를 위하여 AGV가 많이 사용되고 있다. AGV는 주로 정해진 장소 간에 화물을 운송하는 목적으로 사용되며, 매우 고가이다. 산업현장에서 는 AGV 전용 메인제어장치를 자체적으로 PLC를 사용하여 간이적으로 개발하는 경우가 많다. 간이적으로 개발한 AGV는 단순 기능으로만 국한되어 있어 주행 중에 많은 오류를 범한다. 이러한 문제점을 해결하기 위해 전용 AGV 제어가 개발되어 있지만, 이 또한 각 생산 공장의 특징과 구조에 따라서 용도에 맞게 변경이 불가능하다. 본 연구에서는 AGV 주행 흔들림 개선과 공장의 특성에 따라 자유롭게 변경이 가능한 PLC 연동형 AGV 전용 제어 시스템의 설계 방법을 제안하였으며, 실험을 통하여 여러 가지 주행 성능을 분석한 결과, 제안된 AGV 시스템이 운영상에 효율적임을 알 수 있었다. AGV is most often used for the automation of freight transportation in the manufacturing industry or the distribution warehouses. AGV is fairly costly and used in goods transport between the determined location. The main control device dedicated to AGV is often autonomously developed for simplification using PLC in industry field. However, AGV developed for simplification makes many errors in traffic because it is developed for the limited simple function. Control device dedicated to AGV has been developed to solve the problem but it is almost impossible to revise the device according to the character and structure of every manufacturing industry. The purpose of this study is to propose the design method of the control system interlocked with PLC and dedicated to AGV. The control system should be to improve the problem of traffic rolling and possible to be revised according to the character of factory. It is apparent the proposed AGV system is efficient in operation in the result of several traffic performance analysis through the tests.

회전 역진자의 위치제어를 위한 적응 퍼지-신경망 제어기의 설계

탁한호,추연규 진주산업대학교 농업기술연구소 2000 農業技術硏究所報 Vol.13 No.-

In this paper, an position control method using adaptive fuzzy-neural controller(AFNC) is proposed for modeling of nonlinear complex systems. The proposed adaptive fuzzy-neural controller implements system structure and parameter identification using the intelligent schemes together with optimization theory, linguistic fuzzy implication rules, and neural networks from input and output data of processes. Inference type for this adaptive fuzzy-neural controller is presented as simplified inference. To obtain optimal model, the learning rates and momentum coefficients of adaptive fuzzy-neural controller are tuned automatically using improved modified complex method and modified leaning algorithm. For the purpose of its application to nonlinear processes, data for rotating inverted pendulum system ar used for the purpose of evaluating the performance of the proposed adaptive fuzzy-neural controller. The results show that the proposed method can produce the intelligence model with higher accuracy than other works achieved previously.

적응 극배치기법을 이용한 비최소 위상시스템 제어

탁한호,박순규 진주산업대학교 1995 산업과학기술연구소보 Vol.- No.2

본 논문에서는 미지의 시불변 단일 입출력 이산시간 공정을 제어하기 위하여 보조 신호들을 사용한 직접 적응극배치제어기를 설계하였다. 또한 직접 적응극배치제어 알고리즘에서 모든 과거의 입출력 데이타들이 지수함수적으로 가중되도륵 감도함수 여파기를 선정하였다. 직접 적응극배치제어기는 Bezout 항등식을 도입한 직접 적응극배치제어기를 일반화하기 위하여 보조 신호들을 도입하였는데, 그 결과 선형추정 구조를 갖는 선형방정식 오차모델을 구할 수 있었고, 제어대상 공정을 비최소 위상공정의 제어가 가능함을 확인하였으며, 제안된 직접 적응극배치제어기의 응답특성이 개선되었음을 알 수 있었다. In this paper, a direct adaptive pole placement controller for an unknow linear time invariant signal input-output nonminmum phase system is proposed. To design this direct adaptive pole placement controller, the auxiliary signals are introduced. Consequently, a linear equation error model is formulated for estimating both the controller parameters and the additional auxiliary parameters. To estimate the controller parameters and the additional auxiliary parameters, the exponentially weighted least squares algorithm is implemented, and a method of selecting the characteristic polynomials of the sensitivity function filters is proposed. A series of simulations for a nonminmum phase system is presented to illustrate some features of both the parameter estimation and the output response of this adaptive pole placement controller.

GMDP 신경모델을 이용한 비선형시스템의 학습속도 개선에 관한 연구

卓漢浩,金昌圭,金炫悳 진주산업대학교 1995 論文集 Vol.34 No.-

The traditional MLPs have practical difficulties: for examle, a large number of units is required and the learning speed is slow. This paper presents a noval neuron model, called Generalized Multi Dendrite Product(GMDP) unit. Multilayer feedforward neural networks with GMDP units are show to be capable of realizing higher-order neural networks. The standard backpropagation learning rule is extended to this neural network. Simulation rules show that single layer GMDP networks provide an efficient model for solving general problems on function approximation.