• Transactions of the Korean Society of Mechanical Engineers
• /
• v.9 no.4
• /
• pp.509-517
• /
• 1985

작업 공간내에서 원하는 속도와 가속도로, 주어진 경로를 따라 이동하는 k차원 좌표계를 구성하고, 메니퓰레이터의 운동 방정식을 이 좌표계로 변환하여 운동 경로에 대한 선형화 식을 구하였다. 이 시스템의 입력을 변위와 속도의 함수로 정의한 후 안정성을 고려하여 이득을 결정하여 비례-적분제어 시스템을 구성하였다. 이와 같이 구한 적응 제어 알고리즘은 메니퓰레이터의 동적 특성에 대한 정확한 지식을 요하지 않고 또 계산이 간단하여 실시간 응용이 가능하다. 예로서 3차원 공간상의 반경 10cm의 원궤도에 적용하였을 때 최대 오차는 대략 1mm이었으며 상황 변화에 무감각함을 보였다.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.14 no.3
• /
• pp.513-525
• /
• 1990

This paper presents Adaptive Robust Servocontrol(ARSC) scheme, which is an explicit(or indirect) pole-assignment adaptive algorithm with the property of "robustness". It guarantees asymptotic regulation and tracking in the presence of finite parameter perturbations of the unknown plant(or process) model. The controller structure is obtained by transforming a robust control theory into an adaptive control version. This controller structure is combined with the model estimation algorithm which includes a dead-zone for bounded noise. It is proved theoretically that this combination of control and identification is globally convergent and stable. It is also shown, through a real-time simulation study, that the desired closed-loop poles of the augmented system can be assigned directly, and that the adjustment mechanism of the scheme tunes the controller parameters according to the assigned closed-loop poles.oop poles.

• Journal of the Korean Society for Precision Engineering
• /
• v.9 no.3
• /
• pp.73-82
• /
• 1992

An industrial robot, installed real manufacturing processes an element of the system autmation, can be considered as an uncertain system due to dynamic uncertainties in inertial parameters and varying payloads. Most difficuties in controlling a robot manipulator are caused by the fact that the dynamic equations describing the motions of the manipulator are inherently nonlinear and heavily coupled effects between joints and associated links. Existing robot conrol systems have constant predefined gains and do not cover the complex dynamic interactions between manipulator joints. As a result, the manipulator is severly limited in range of application, speed of operation and variation of payload. The proposed controller is operated by adjusting its gains based on the response of the manipulator in such a way that the manipulator closely matches the reference model trajectories defined by the desinger. The proposed manipulator studied has two loops, an inner loop of model reference adaptive controller and an outer loop of state feedback controller with integral action to guarantee the stability of the adaptive scheme. This adaptation algorithm is based on the hyperstailiy approach with an improved Lyapunov function. The coupling among joints and the nonlinearity in the dynamic equation are explicitly considered. The designed manipulator controller shows good tracking performance in practical working environment, various load variations and parameter uncertainties.