Title :
Controller estimation for the adaptive control of robotic manipulators
Author :
Guo, Lin ; Angeles, Jorge
Author_Institution :
CSIRO, Preston, Vic., Australia
fDate :
6/1/1989 12:00:00 AM
Abstract :
An adaptive control method for linearized models of robotic manipulators is presented. First, the highly coupled nonlinear dynamical equations of robotic manipulators are linearized. Based on the linearized equations, the controllers are designed so as to render the feedback model a prescribed second-order stationary system, under the assumption that the deviations of position, velocity, and acceleration from their nominal values are small. A recursive least-square identification scheme is used to perform the online parameter estimation for the feedback gains of the controllers, instead of estimating the inertial parameters of the robotic manipulator themselves. The arising algorithm is much simpler and the required computations are reduced, when compared with a similar scheme (see C.S.G. Lee and M.I. Chung, 1984). Simulation of a PUMA 600 robotic manipulator under errors in the inertial parameters between 5% and 15% shows that the position and velocity errors are within reasonable limits
Keywords :
adaptive control; dynamics; feedback; least squares approximations; parameter estimation; robots; PUMA 600 robotic manipulator; adaptive control; feedback model; linearized models; nonlinear dynamical equations; online parameter estimation; recursive least-square identification; robotic manipulators; second-order stationary system; Acceleration; Adaptive control; Couplings; Linear feedback control systems; Manipulator dynamics; Nonlinear equations; Parameter estimation; Performance gain; Robots; Velocity control;
Journal_Title :
Robotics and Automation, IEEE Transactions on