• 한국지능시스템학회논문지
• /
• 제7권3호
• /
• pp.3-11
• /
• 1997

비선형 동적 시스템을 제어하기에 적합한 귀한 신경망에 대한 주요한 연구중의 하나는 안정적인 학습 알고리듬을 개발하는 것이다. 본 논문에서는 진화 연산 알고리듬을 이용한 안정작인 귀환 신경망의 학습 알고리듬을 개발한다. 또한 개발한 학습 알고리듬을 사용한 귀환 신경망을 전형적인 비선형 동적 시스템인 로보트 매니퓰레이터의 제어 시스템에 적용하고 개발한 제어 알고리듬의 효용성을 입증한다.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 1995년도 춘계학술대회 논문집
• /
• pp.262-266
• /
• 1995

This paper presents a new adaptive-neuro control scheme to control the velocity and position of SCARA robot with parameter uncertainties. The adaptive control of linear system found wiedly in many areas of control application. While techniques for the adaptive control of linear systems have been well-established in the literature, there are a few corresponding techniques for nonlinear systems. In this paper an attempt is made to present a newcontrol scheme for theadaptive control of ponlinear robot based on a feedforward neural network. The proposed approach incorporates a neuro controller used within a reinforcement learning framework, which reduces the problem to one of learning a stochastic approximation of an unknown average error surface Emphasis is focused on the fact that the adaptive-neuro controoler dose not need any input/output information about the controlled system. The simulation result illustrates the effectiveness of the proposed adaptive-neuro control scheme.

• 한국정밀공학회지
• /
• 제6권1호
• /
• pp.52-74
• /
• 1989

In general, the state of system can be effected by external noise and observed only through a noisy channel. Therefore we use the estimation technigue for the information of state of the system effected by noise. There are many filters such as kalman-Buchy filter, kalman filter, Extended Kalman filter algorithm, cononlinear, extended Kalman filter algorithm to the estimation of parameters is very useful and has a long history. Also a considerable number of applications of this method has been reported. In this paper, the robot control system is treated in stochastic optimal control because of the robots doing a complicated and accurate task in inapproate environment. We have conclusion that error covariance is converged and the stability of filtering is obtained.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1988년도 전기.전자공학 학술대회 논문집
• /
• pp.273-277
• /
• 1988

This paper is to show design of robot manipulator which has 3-link using DC Motor and realization of control algorithm with IBM - XT Micro-computer connected. Gentral algorithm is applicated by position and pass control using point-to-point method. At first, this paper computes required angles on each joint in order to search desired position or path, and uses a voltage control with feedback from output of encoder and tachometer in real time. The application of control algorithm on position, velocity and force for each joint of manipulator by using self-tuning control is left for next study.

• 대한전자공학회논문지
• /
• 제25권2호
• /
• pp.144-151
• /
• 1988

Between any two points in the Cartesian space, the straight motion has many distinct advantages in path/trajectory plannings, specially in a crowded workspace. But to achieve this motion is a nontrivial task. One standard way to approximate this motion is to plae enough intermediate points along the desired path and linearly interpolate two adjacent intermediate points in the joint space. In this approach, however, the determination of the minimum number of intermediate points is very important from several aspects. A scheme that can effectively accomplish this purpose is established in this paper. This scheme is based on several search methods. The results are demonstrated using the PUMA 560 series manipulator.

• 대한전자공학회논문지
• /
• 제25권2호
• /
• pp.152-161
• /
• 1988

In this paper, a direct method for obtaining the trajectory set points is investigated in discrete-time, which is different from the other conventional schemes. We consider the tracking of a straight line path, where the trajectory set points for manipulator control are determined exactly on the straight line path. For the purpose of the munimum-time operation of manipulators, the problem is formulated as a maximization of the Cartesian distance between two consecutive servo time instants. The maximization is subject to the smoothness and torque constraints. Several algorithms are developed and utilized to maximize the Cartesian distance. The proposed approach has been simulated on a VAX-11/780 computer to verify its performance.

• 대한전자공학회논문지
• /
• 제27권9호
• /
• pp.1364-1371
• /
• 1990

We propose a time-optimal path planning scheme for the robot manipulator using Hopfield neural network. The time-optimal path planning, which can allow the robot system to perform the demanded tasks with a minimum execution time, may be of consequence to improve the productivity. But most of the methods proposed till now suffers from a significant computational burden and thus limits the on-line application. One way to avoid such a difficulty is to apply the neural networke technique, which can allow the parallel computation, to the minimum time problem. This paper proposes an approach for solving the time-optimal path planning by using Hopfield neural network. The effectiveness of the proposed method is demonstrarted using a PUMA 560 manipulator.

• 한국지능시스템학회논문지
• /
• 제6권2호
• /
• pp.111-119
• /
• 1996

In this paper, we propose time-optimal trajectory planning algorithms for cooperative multi-robot manipulators system considering optimal load distribution. Internal forces essentially effect on time optimal trajectory planning and if they are comitted, the time optimal scheme is not no longer true. Therefore, we try to find the internal force factors of cooperative robot manipulators system in a time-optimal aspect. In this approach, a specific generalized inverse is used and is fuzzified for the purpose. In this optimal method, the fuzzy logic concept is used and selected for diminishing computation time, for finding the load distribution factors.

• 전자공학회논문지B
• /
• 제28B권11호
• /
• pp.934-940
• /
• 1991

The multi-layer neural network that has broadly been utilized in designing the controller of robot manipulator possesses the desirable characteristics of learning capacity, by which the uncertain variation of the dynamic parameters of robot can be handled adaptively, and parallel distributed processing that makes it possible to control on real-time. However the error back propagation algorithm that has been utilized popularly in the learning of the multi-layer neural network has the problem of its slow convergencs speed. In this paper, an approach to improve the convergence speed is proposed using fuzzy logic that can effectively handle the uncertain and fuzzy informations by linguistic level. The effectiveness of the proposed algorithm is demonstrated by computer simulation of PUMA 560 robot manipulator.

• 대한산업공학회지
• /
• 제22권3호
• /
• pp.499-515
• /
• 1996

The basic function of on industrial robot system is to move objects in the workspace fast and accurately. One difficulty in performing this function is that the path of robot should be programmed to avoid the collision with obstacles, that is, tools, or facilities. This path planning requires much off-line programming time. In this study, a SOM technique to find the collision-free path of robot in real time is developed. That is, the collision-free map is obtained through SOM learning and a collision-free path is found using the map in real time during the robot operation. A learning procedure to obtain the map and an algorithm to find a short path using the map is developed and simulated. Finally, a path smoothing method to stabilize the motion of robot is suggested.