Modeling and control of aerial robotics platforms for environment interaction: comparative analysis of Flying Hand and Unmanned Aerial Manipulator

The transition of Unmanned Aerial Vehicles (UAVs) from passive observers to active workers has introduced the complex paradigm of Aerial Physical Interaction (APhI). However, physical contact with the environment generates strong, time-varying dynamic coupling that severely compromises flight stability. A critical open problem in this field is determining the optimal architectural design for interaction tasks: the "Flying Hand" (FH), featuring a tilted-propeller hexarotor base with a rigidly attached tool, versus the "Unmanned Aerial Manipulator" (UAM), a coplanar hexarotor equipped with an articulated robotic arm. This thesis systematically investigates and compares these two configurations. To accurately capture the complex coupling torques, robust mathematical models were developed using the Newton-Euler formalism, specifically exploiting the Recursive Newton-Euler algorithm for the UAM. The problem was addressed by designing and simulating a rigorous benchmark scenario in MATLAB/Simulink, where the aerial robots transition from free-flight trajectory tracking to applying a constant force on a target after autonomously maneuvering toward said target. By evaluating different model-based controllers, the systems' dynamic responses were quantitatively analyzed. The main finding is that the UAM described is mechanically superior to the FH for operations requiring contact with the external environment, as its actuated arm absorbs interaction forces and torques, thereby ensuring greater stability for the overall system during the task.