Optimal Interaction-friendly Trajectory Generation of a Pneumatic Soft Continuum Manipulator

摘要

The technology of soft continuum robots represents an advance in the field of robotics to benefit a wide range of industries such as healthcare, manufacturing or environmental exploration. Due to their compliant continuous structure, they are adaptive to environmental conditions and thus beneficial for physical human-robot interaction (pHRI). The use of pneumatic actuation offers crucial safety benefits for pHRI due to its direct drive characteristics and its inherent compliance by the compressibility of air. In this paper, an interaction-friendly, optimization-based method for offline trajectory generation of pneumatic soft continuum manipulators for allowing undesired and desired interactions in the context of pHRI is derived and validated experimentally on a handling task performed by the bionic soft arm. The interaction-friendly criteria are mainly described by the manipulator’s stiffness and the detectability of external forces, which are considered in a two-step optimal control problem for the free movement and the contact approach of the bionic soft arm. The resulting optimal trajectories of the joint angles and joint stiffnesses are compared with time-optimal trajectories and discussed. From the measurements, it can be concluded that the presented method exploits the redundancies of the differential pressure actuation and the properties of the manipulator to achieve interaction-friendly trajectories.