Adaptive Control of Robotic Manipulators Using an Extended Kalman Filter

Abstract

This paper presents a new adaptive motion control scheme for robotic manipulators. This is an adaptive computed torque method (CTM) that requires only position measurements. These measurements and the input torques are used in an extended Kalman filter (EKF) to estimate the inertial parameters of the full non-linear robot model as well as the joint positions and velocities. These estimates are used by the CTM to generate the input torques. The theory behind Kalman filtering provides clear guide-lines on the selection of the design parameters for the controller when noise is present. Simulation results illustrate the performance of this scheme and demonstrate its noise rejection properties.