Physical Human-Robot Interaction (pHRI) through Admittance Control of Dynamic Movement Primitives in Sit-to-Stand Assistance Robot

Abstract

Physically assistance robots have been conceptualized to compensate or augment the human musculoskeletal function. However, due to the concerns of safety and effectiveness for physical human-robot interaction (pHRI) in such robots, compliant joints are preferred over the rigid joints. This paper illustrates the implementation of admittance control in a sit-to-stand (STS) assistance robot. The 3-degrees of freedom (dof) robot comprises of ball screw-based linear actuators that are arranged in a parallel configuration. While the actuation system is preferable for strength and performance, the non-backdrivable characteristic corroborates the rigidity of the joint, making it unfavorable for human-robot interaction operation. To enhance compliance, force sensor-based admittance control system is implemented. Regarding motion planning, the trajectory were modeled as Dynamic Movement Primitives (DMP), which facilitates the implementation of admittance control. The proposed model is implemented in the robot prototype and validated by illustrating the force input and the motion output.