• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2000년도 제15차 학술회의논문집
• /
• pp.431-431
• /
• 2000

The modeling of 5-bar linkage robot manipulator dynamics by means of a mathematical and neural architecture is presented. Such a model is applicable to the design of a feedforward controller or adjustment of controller parameters. The inverse model consists of two parts: a mathematical part and a compensation part. In the mathematical part, the subsystems of a 5-bar linkage robot manipulator are constructed by applying Kawato's Feedback-Error-Learning method, and trained by given training data. In the compensation part, MLP backpropagation algorithm is used to compensate the unmodeled dynamics. The forward model is realized from the inverse model using the inverse of inertia matrix and the compensation torque is decoupled in the input torque of the forward model. This scheme can use tile mathematical knowledge of the robot manipulator and analogize the robot characteristics. It is shown that the model is reasonable to be used for design and initial gain tuning of a controller.

• 한국지하수토양환경학회:학술대회논문집
• /
• 한국지하수토양환경학회 2006년도 총회 및 춘계학술발표회
• /
• pp.391-393
• /
• 2006

In this study, the performance of artificial neural network (ANN) models for estimating spatial distribution of water quality was evaluated using electric conductivity (EC) values in landfill site. For the ANN model development, feedforward neural networks and backpropagation algorithm with gradient descent method were used. In Test 1, the interpolation ability of the ANN model was evaluated. Results of the ANN model were more precise than those of the Kriging model. In Test 2, spatial distributions of EC values were predicted using precipitation data. Results seemed to be reasonable, however, they showed a limitation of ANN models in extrapolations.

• 한국공작기계학회:학술대회논문집
• /
• 한국공작기계학회 2000년도 추계학술대회논문집 - 한국공작기계학회
• /
• pp.108-112
• /
• 2000

Neural networks are used in the framework of sensorbased tracking control of robot manipulators. They learn by practice movements the relationship between PSD (an analog Position Sensitive Detector) sensor readings for target positions and the joint commands to reach them. Using this configuration, the system can track or follow a moving or stationary object in real time. Furthermore, an efficient neural network architecture has been developed for real time learning. This network uses multiple sets of simple backpropagation networks one of which is selected according to which division (corresponding to a cluster of the self-organizing feature map) in data space the current input data belongs to. This lends itself to a very training and processing implementation required for real time control.

• 한국산학기술학회:학술대회논문집
• /
• 한국산학기술학회 2006년도 춘계학술발표논문집
• /
• pp.311-315
• /
• 2006

"Dynamic Neural Unit"(DNU) based upon the topology of a reverberating circuit in a neuronal pool of the central nervous system. In this thesis, we present a genetic DNU-control scheme for unknown nonlinear systems. Our methodis different from those using supervised learning algorithms, such as the backpropagation (BP) algorithm, that needs training information in each step. The contributions of this thesis are the new approach to constructing neural network architecture and its trainin

• 조명전기설비학회논문지
• /
• 제14권2호
• /
• pp.8-17
• /
• 2000

본 연구에서는 부분방전의 인식을 위한 두가지 접근법을 제안하였다. 첫 번째로는 백프로파게이션 알고리즘을 적용한 신경회로망이구 두 번째로는 두 연산자 백터간의 사이각 계산이다. 부분방전신호는 IEC(b), 침대평판 및 CIGRE method II 등 3가지 전극계로부터 각각 검출되었다. 신경회로망과 벡터의 사이각을 이용한 방법 모두 이리 학습된 패턴에 대해서는 양호한 인식능력을 보였다. 그려고 사용되는 연산자의 수가 미학습패턴의 인식능력에 큰 영향올 미쳤다.

• 조명전기설비학회논문지
• /
• 제12권3호
• /
• pp.7-13
• /
• 1998

본 논문에서는 시스템의 동특성이 변화 가능한 구간에서 임의의 생플올 추출하여 신경회로망을 학습시킴으로써 시스템 동특성에 따라 Pill 제어기가 자동동조하는 방법을 제시하였다. 신경회로망을 학습시키기 위해 역전파 학습 알고리즘올 사용하였고, 교사치로는 규칙기반에 의해서 얻어진 매개변수를 이용하였다. 부하 변화시 시스템의 동특성에 맞는 Pill 제어기가 자동동조됨을 시뮬레이션에 의해 확인하였다.

• 대한전기학회논문지
• /
• 제39권9호
• /
• pp.985-996
• /
• 1990

An inverse kinematic solution of a robot manipulator using multilayer perceptrons is proposed. Neural networks allow the solution of some complex nonlinear equations such as the inverse kinematics of a robot manipulator without the need for its model. However, the back-propagation (BP) learning rule for multilayer perceptrons has the major limitation of being too slow in learning to be practical. In this paper, a new algorithm named Dynamically Reconfiguring BP is proposed to improve its learning speed. It uses a modified version of Kohonen's Self-Organizing Feature Map (SOFM) to partition the input space and for each input point, select a subset of the hidden processing elements or neurons. A subset of the original network results from these selected neuron which learns the desired mapping for this small input region. It is this selective property that accelerates convergence as well as enhances resolution. This network was used to learn the parity function and further, to solve the inverse kinematic problem of a robot manipulator. The results demonstrate faster learning than the BP network.

• 대한전기학회논문지:전력기술부문A
• /
• 제48권7호
• /
• pp.815-823
• /
• 1999

This study considers an implementation of artificial neural networks to the receding horizon optimal control and is applications to power systems. The Generalized Backpropagation-Through-Time (GBTT) algorithm is presented to deal with a quadratic cost function defined in a finite-time horizon. A decentralized approach is used to control the complex global system with simpler local controllers that need only local information. A Neural network based Receding horizon Optimal Control (NROC) 1aw is derived for the local nonlinear systems. The proposed NROC scheme is implemented with two artificial neural networks, Identification Neural Network (IDNN) and Optimal Control Neural Network (OCNN). The proposed NROC is applied to a power system to improve the damping of the low-frequency oscillation. The simulation results show that the NROC based power system stabilizer performs well with good damping for different loading conditions and fault types.

• 한국컴퓨터정보학회논문지
• /
• 제1권1호
• /
• pp.83-94
• /
• 1996

ANOVA 검정에서 검정통계량은 단일 또는 이중 비중심F분포를 따르며 비중심F분포는 일반적인 선형 가설 검정에서 검정함수 계산에 적용되고 있다. 본 논문에서는 단일 비중심F분포의 누적함수 계산에 신경망이론을 적용하였다. 신경망 구조는 다층 퍼셉트론이며 학습과정은 역전과 학습알고리즘이다. 신경망이론에 의하여 계산한 결과와 Patnaik 이 제시한 확률값을 비교하여 제시하였다.

• 한국시뮬레이션학회논문지
• /
• 제9권3호
• /
• pp.77-89
• /
• 2000

This paper suggests a multi-attribute dispatching rule for an automated guided vehicle(AGV). The attributes to be considered are the number of queues in outgoing buffers of workstations, distance between an idle AGV and a workstation with a job waiting for the service of vehicle, and the number of queues in input buffers of the destination workstation of a job. The suggested rule is based on the simple additive weighting method using a normalized score for each attribute. A neural network approach is applied to obtain an appropriate weight vector of attributes based on the current status of the manufacturing system. Backpropagation algorithm is used to train the neural network model. The proposed dispatching rules and some single attribute rules are compared and analyzed by simulation technique. A number of simulation runs are executed under different experimental conditions to compare the several performance measures of the suggested rules and some existing single attribute dispatching rules each other.