Compensation of Load Dynamics for Admittance Controlled Interactive Industrial Robots Using a Quaternion-Based Kalman Filter

The paper describes a control architecture for industrial robotic applications allowing human/robot interactions, using an admittance control scheme and direct sensing of the human inputs. The aim of the proposed scheme is to support the operator of an industrial robot, equipped with a force/torque (F/T) sensor on the end-effector, during human/robot collaboration tasks involving heavy payloads carried by the robot. In these practical applications, the dynamics of the load may significatively affect the measurements of the F/T sensor. Model-based compensation of such dynamic effects requires to compute linear acceleration and angular acceleration/velocity of the load, that in this paper are estimated by means of a quaternion-based Kalman filter and assuming that the only available measurements come from the forward kinematics of the robot. Experimental results demonstrate the feasibility of the approach and its industrial applicability.