COMPUTER GRAPHICS AIDED ADAPTIVE MANIPULATOR CONTROL
Abstract
The pole-placement self-tuning control method is applied to the control of manipulators tin the task coordinate system. This results in desirable uncoupled task-oriented dynamic performance characteristics. The effective dynamics of the end-effector or tool along each position and orientation axis in the task coordinate system is first identified by using an on-line recursive-least-square estimation algorithm. The pole-placement control design method uses the identified model to calculate the pseudo-control input for each position and orientation axis. This pseudo-control input vector at the task level is transformed back to the joint actuators via the transpose of the generalized Jacobian matrix to exert new input torques at the manipulator joints. In the task coordinate, the position degrees of freedom and force degrees of freedom are known to be orthogonal. This allows task level control to be composed of N independent controllers, each based on a task direction error learning mechanism. Thus the task level controller is independent of kinematic configuraion changes, dynamic friction, compliant effects and load variations. The new manipulator control method can also be applied to the control of redundant manipulators. Simulations of both position control and hybrid control using the JPL-Stanford Arm demonstrate the effectiveness of the proposed method. Real-time implementaion of this algorithm for position control of the PUMA 600 industrial manipulator further supports the applicability of this new control method. To enhance the capabilities of a manipulator system, computer graphics techniques for task planning, teaching, simulation and on-line feedback control of a robot manipulator are presented. It is shown that by incorporating human decision-making capabilities in the feedback control loop, many tasks can be executed more easily and more efficiently. A sophisticated 3-D graphic device communicating with a VAX 11/780 minicomputer is used to demonstrate the feasibility of these concepts through simulations of a PUMA 600 industrial manipulator.
Degree
Ph.D.
Subject Area
Mechanical engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.