• 제어로봇시스템학회:학술대회논문집
• /
• 1996.10b
• /
• pp.698-701
• /
• 1996

A novel neural network control scheme is proposed to identify the inverse dynamic model of robot manipulator and to compensate for uncertainties in robot dynamics. The proposed controller is called reference compensation technique(RCT) by compensating at reference input trajectory. The proposed RCT scheme has many benefits due to the differences in compensating position and learning algorithm. Since the compensation is done outside the plant it can be applied to many control systems without modifying the inside controller. It performs well with low controller gain because the operating range of input values is small and the output of the neural network controller is amplified through the controller gain. The back-propagation algorithm is used to train and simulations of three link robot manipulator are carried out to prove the proposed controller's performances.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.129.5-129
• /
• 2001

When a robot manipulator performs some task like grinding or assembling, not only the position control but also the force control of the tools connected to the robot must be controlled. But at this time We were received the uncertainty problems of system information for the force control, for example disturbance, senor resolution and measurement noise. Therefore we proposed fuzzy logic control method instead of existing control theory for the robot manipulator control, for example PID control method. In this paper, We proposed hybrid position/force control of robot manipulator using fuzzy logic control method. To show the validity of the proposed fuzzy controller, We compared fuzzy controller with conventional PID controller.

• Journal of the Institute of Convergence Signal Processing
• /
• v.2 no.2
• /
• pp.89-94
• /
• 2001

The position control accuracy of a robot manipulator is significantly deteriorated when a long arm robot is operated at a high speed. This paper presents a very simple sliding mode control which eliminates multiple mode residual vibration in a robot manipulator. The neural network is used to avoid that sliding mode condition is deviated due to the change of system parameter and disturbance. This paper is suggested control system which designed by sliding mode controller using neural network. The effectiveness of proposed scheme is demonstrated through computer simulation.

• The Journal of Korea Robotics Society
• /
• v.11 no.4
• /
• pp.235-241
• /
• 2016

Dual arm manipulators have been developed for the entertainment purpose such as humanoid type or the industrial application such as automatic assembly. Nowadays, there are some issues for applying the dual arm robot system into the various fields. Especially, robots can substitute human and perform the dangerous activity such as search and rescue in the battle field or disaster. In the paper, the dual arm manipulator which can be adapted to the rescue robot with the mobile platform was developed. The kinematic design was proposed for the rescue activity and the required specification was determined through the kinematic analysis and the dynamic analysis in the various conditions. The proposed dual arm manipulator was manufactured based on the vibration analysis result and its performance was proved by the experiment.

• Proceedings of the KIEE Conference
• /
• 1995.11a
• /
• pp.575-578
• /
• 1995

In general, the control of robot manipulator is classified into position control and force control. Position controllers give adequate performance when a manipulator is following a trajectory through space and end-effector has no contact with environment. However for most tasks performed by robot manipulator in industry, contact is made between the end-effector and manipulator's environment, so position control may not suffice. The objective of this study is to control both position of a manipulator and the contact forces generated at the hand by using a conceptually simple control law. Position and force control problem is decoupled into subtasts via taskspace formulation and inverse dynamics. Then, the position controllers are designed for the task space variable which represent tangent motion and the forte controllers are designed for the lash space variables which represent normal force.

• Journal of the Korean Society of Manufacturing Technology Engineers
• /
• v.8 no.5
• /
• pp.76-82
• /
• 1999

During operating given working a robot manipulator makes some problems such as the accumulation of the error or the deviation from the command trajectory. These problems are mainly due to the disturbance noise or unmodeled system parameters. To solve these problems most of robot manipulators equip the controller. But if exact controller gains are not seleced we can't decrease the working efficiency(such as compensation about error or deviation) of the robot manipulator. So in this paper we present the controller gain tuning law by which we can find the controller gain which satisfies the per-formance specification of the robot manipulator during working of the robot. The proposed algorithm is derived from the Laypunov direct method. And by the simulation on the 4-axis SCARA type robot(SAMSUNG SM5 Robot) we guarantee the performance of this algorithm.

• 제어로봇시스템학회:학술대회논문집
• /
• 1990.10a
• /
• pp.596-601
• /
• 1990

When using robot manipulator to carry out autonomous tasks, the positioning accuracy of the robot manipulator relative to a reference coordinate frame is of greate importance. The task program, which is generated by off-line CAD-system and used in actual robot positioning, may cause serious amount of the absolute positioning error of the robot manipulator. In this study, a robot performance test and calibration algorithms are proposed in order to improve the absolute positioning accuracy of the robot end effector. Experiments were also carried out by utilizing the HYUNDAI Robot AE 7601 and KIM2-Tester, a three dimensional measurement system, which is developed in Robotics & Fluid Power Control Lab. at Korea Institute of Science and Technology.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.28 no.5
• /
• pp.539-546
• /
• 2004

Conventional robot manipulators actuated by motors with the speed reducer such as the harmonic drive have weakness in the load capacity, since the speed reducer does not have enough strength. To overcome this, a new type of the robot actuator based on the four-bar-link mechanism driven by the ball screw was proposed and constructed. Also, a new type of a revolute-jointed robot manipulator composed of the developed actuators was developed. The base axis is actuated by the motor with the conventional speed reducer, but the other axes are actuated by the proposed actuators. The kinematics and dynamics of the robot were analyzed, and the performance test of the robot was made. Through the test results, the performance of superior load capacity versus the robot weight is shown.

• 제어로봇시스템학회:학술대회논문집
• /
• 1997.10a
• /
• pp.548-551
• /
• 1997

Performance of force tracking impedance control of robot manipulators is degraded by the uncertainties in the robot and environment dynamic model. The purpose of this paper is to improve the controller robustness by applying neural network. Neural networks are designed to learn the uncertainties in robot and environment model for compensating the uncertainties. The proposed scheme is verified through the simulation of 20DOF robot manipulator.

• Journal of Advanced Marine Engineering and Technology
• /
• v.27 no.2
• /
• pp.238-245
• /
• 2003

To overcome the weakness in the load capacity of conventional robot manipulators actuated by motors with the speed reducer such as the harmonic driver, we proposed a new closed-chain type of the robot actuator which is composed of the four-bar-link mechanism driven by the ball screw. The robot manipulator is revolute-jointed and composed of four axes. The base axis is actuated by the lineal actuator such as the ball screw, and the others are actuated by the proposed actuator. We analyzed the mechanism of the actuators of the robot joints, and developed the dynamics model. The dynamics are expressed in the joint coordinates and then they are mapped into the sliding coordinates of the ball screw. We performed fundamental tests on the structure of the robot.