A Hybrid Mapping Method with Position and Stiffness for Manipulator Teleoperation

Abstract

Transferring versatile skills of human behavior to teleoperate manipulators to execute tasks with large uncertainties is challenging in robotics. This paper proposes a hybrid mapping method with position and stiffness for manipulator teleoperation through the exoskeleton device combining with the surface electromyography (sEMG) sensors. Firstly, according to the redefinition of robot workspace, the fixed scale mapping in free space and virtual impedance mapping in fine space are presented for position teleoperation. Secondly, the stiffness at the human arm endpoint is predicted and classified into three levels based on the K nearest neighbor (KNN) and XGBoost, and the stiffness mapping method is utilized to regulate the stiffness behavior of manipulator. Finally, the proposed method is demonstrated in three complementary experiments, namely the trajectory tracking in free space, the obstacle avoidance in fine space and the human robot interaction in contact space, which illustrate the effectiveness of the method.