Optimization-Based Compliant Controller for Physical Human-Aerial Manipulator Interaction

The use of aerial manipulators in mixed human-robot environments, necessitates the capacity for safe interactions between them. This work is concerned with enabling human-aided navigation of unmanned aerial manipulators, allowing human operators to effectively take over the path planning task of an aerial manipulator platform, by exerting appropriate forces on its end-effector. A guaranteed-compliance model-based optimization controller is developed for the system's articulated arm, ensuring that the arm can comply with forces and moments on its end-effector, while using control barrier functions (CBFs) to maintain non-singular configurations at all times. Another optimization-based controller is designed for the aerial vehicle, appropriately interpreting the robot arm end-effector motions in order to comply with human-induced intended poses, with CBFs included to ensure measured forces remain bounded, for additional safety. Experimental studies are included, showcasing the capacity of the presented control framework in enabling human-guided navigation for autonomous aerial manipulators.