Contact Force Distribution Predictive Control System in Wearable Robot with Tactile Sensor

Abstract

Ensuring safety for the wearer is of utmost importance for wearable robots. For safe assistance of complex body parts such as the torso, it is important to measure and control the contact force: the force applied to the human body from the robot. We have been studying the robot enabling to measure contact force distribution from tactile sensors on the robot surface as pressure distribution. In order to take safe assistance, the robot is controlled to reduce superfluous force by direct feedback control of the contact force in wearable robots. In this paper, a new control method is proposed, in which predictive control of contact force distribution is applied. The method predicts the contact force distribution of the next step from the current link positions and measured contact force distribution. The robot controls his actuators to minimize the predictive errors with the optimum contact force distribution. The proposed method is compared with the conventional control based on inverse kinematics, being verified the effectiveness.