Parameter Identification of an Unknown Object in Human-Robot Collaborative Manipulation

Abstract

In human-robot collaborative manipulation of an object, if the robot knows the intention of the human, the efficiency of the work would greatly increase. For the robot to know of the human intention, it should have the information of the force applied by the human, which can be more accurately if it can estimate the inertial and dimensional parameters online. However, the force applied by the human will disturb the parameter identification process. This paper presents a strategy to identify the inertial and dimensional parameters of an unknown object online for physical human-robot interactions. Extended Kalman filter is used for identification under the assumption that the force applied by the human is an unknown external disturbance. This approach was evaluated in simulations of physical human-robot object manipulation task.