Robust disturbance estimation for human–robotic comanipulation†

External forces affect the dynamics of load-carrying robot devices. The knowledge of such disturbances is generally needed for control purposes. However, direct disturbance measurement using force sensors is not always possible. This paper introduces a force estimator for force-sensor-less robotic manipulators. The algorithm is based on the knowledge of the dynamics of the robotic device, whereas mass of the load is typically unknown. Using this algorithm, low-frequency external forces can be estimated robustly even for quasi-statically time-varying and uncertain loads. Experiments validate the proposed strategy in practice. Moreover, the applicability of the estimation algorithm is further illustrated by using it in a human–robot comanipulation setup in which the robot is providing additional coordinated forcing to alleviate human effort needed to manipulate the robot.