Task-space Control of an Unmanned Helicopter Equipped with a 7-DoF Robot Arm

Kurzfassung

Aerial manipulation is generally defined as the grasping, transporting, positioning, and assembly or disassembly of mechanical parts and objects, performed by UAV with a robotic arm. Such types of systems can for instance be used for inspection and maintenance tasks in areas which are inaccessible or potentially hazardous for humans. A dynamic aerial vehicle equipped with a robotic manipulator is a challenging control problem due to the high dynamical coupling between both subsystems. In order to achieve highly dynamic behaviors while enabling precise task-space tracking of the manipulator, this coupling has to be taken into account in the control design. Starting from input-output linearization it is discovered that many outputs are not differentially flat and result in an unstable internal dynamics. By applying a passive decomposition to the system, an interesting underlying structure of the underactuated nonlinear system is revealed, which facilitates the control design. Based on this transformation a control law is proposed, which guarantees task-space (e.g. end-effector) reference tracking and position trajectory control of the overall system. The proposed approach is verified in numerical simulation. In order to analyze if the model based control law is suitable for practical implementation, a more detailed dynamics including unmodeled effects is used in the simulation.