Content area
Abstract
This dissertation presents the implementation and theoretical analysis of four methods of repetitive control used to eliminate known and measurable periodic tracking error in direct drive robot manipulators. The first method investigates a discrete time repetitive control approach which is applied throughout the entire trajectory of a robot during repetitive tasks. Simulation and experimental results show that the repetitive controller in conjunction with a simple Proportional-Derivative (PD) control law assures good tracking performance when the desired trajectory is periodic and its period is known.
A second method is a segmented approach to repetitive control. Unlike the previous design which applies repetitive control throughout the entire repetitive path of the robot, the segmented approach applies repetitive control only to specific segments of the path. The stability of the segmented repetitive control system is presented. Simulation and experimental results of a two link direct drive SCARA-type robot arm under segmented repetitive control are also presented.
The third method in this thesis investigates Cartesian space repetitive control. In Cartesian space repetitive control the feedback signal is acquired from the end effector of the manipulator, in contrast to conventional joint control which utilizes joint oriented sensors such as shaft encoders, or resolvers. Although conventional joint level repetitive control significantly improves the performance of the robot in the presence of dynamic model mismatches and disturbances, Cartesian space repetitive control is necessary if kinematic modelling error is to be eliminated. An X-Y digitizer is used as an End Effector Cartesian Coordinate Sensor (EECCS). The Cartesian space repetitive controller is shown to be successful in eliminating the repetitive error in both simulation and implementation.
An implementation of Plug-In Repetitive Control on the direct drive axis of a prototype GMF A500 robot is also treated in this thesis to illustrate the possible use of repetitive control in an industrial environment. Plug-In Repetitive Control is developed and implemented using an IBM AT. A multirate sampling algorithm is designed to reduce the memory requirements of repetitive control. Experimental results show that the tracking error is smoothly absorbed by the Plug-In repetitive controller in a few cycles.
A two axis direct drive SCARA robot utilizing NSK Megatorque direct drive motors is used to test the algorithms developed in this dissertation.
