REPETITIVE CONTROL OF ROBOTIC MANIPULATORS IN OPERATIONAL SPACE: A NEURAL NETWORK-BASED APPROACH

Abstract

This work tackles the control problem for robotic manipulators with kinematic and dynamical uncertainties where the end-effector robot is required to perform repetitive tasks. Specifically, a neural network-based estimator and an adaptive component have been fused with a repetitive learning controller-based update rule to compensate for the uncertainties in the robot dynamics and parametrically uncertain kinematics. The closed-loop system stability and tracking of periodic desired operational space position vector are ensured via Lyapunov-type analysis. Experiment results obtained from a planar robotic manipulator are presented to demonstrate the feasibility of the proposed control methodology.