異種ロボットチームのためのVTOL-UAVドッキングシステム

In the current trend for the field of robotic exploration, the heterogeneous robot teams have been acquiring great relevance because they offer significant advantages for exploration and navigation missions in unknown environments. Notably, the robot teams with aerial and ground platforms have superior efficiency due to their mixed abilities and shared sensor resources. Within this classification, the robot teams that use vertical take-off and landing (VTOL) aerial platforms and robust ground robots have significant benefits because they can see the same environment from different perspectives, extracting valuable information and adding great redundancy to the measurement process. This type of robot team is ideal for applications that require great precision and efficiency and to explore unknown and dangerous environments. However, the VTOL unmanned aerial vehicles (VTOL-UAV) have limitations due to their high energy consumption and reduced payload capacity, restricting the exploration abilities of the complete robot team. When the VTOL and ground robot platforms are deployed together as a team, at some point in the mission, the aircraft needs to return to the base for landing and battery replenishment to continue with the exploration. This task restricts the team to explore the environment in-depth and to work for longer durations. As a solution to this problem, it is presented in this work an autonomous system for the docking of a VTOL-UAV with the mobile manipulator of the robot team. The proposed system contains a central measurement-actuation unit that consists of a robot manipulator mounted in a mobile platform with a visual sensor configured as an eye-in-hand device. The extracted visual information is used to calculate the necessary arm movements to execute a stable UAV tracking to achieve air-ground robot contact. It is presented with a complete control scheme that merges predictive and optimization approaches to calculate precise commands for both robot platforms. The controller workspace is designed to include environmental and structural constraints for safe air-ground contact. The docking strategy allows the multi-robot team to economize and potentially recover aircraft energy and also enables the robust ground platform to protect and store the aircraft facilitating to continue the mission for a longer duration. The on-site autonomous docking can be initiated multiple times, allowing the team to explore larger areas and complete complex and long-term missions. The main focus of this thesis is to create a robust system capable of performing a safe and repeatable docking task within complex environments, enhancing the robot team and its usability. The proposed method is presented with the corresponding theoretical development and complete hardware and control design. The validity of the system is tested with simulation and experiments in a real environment. The collected data for simulation and the real environment is represented in graphical form, and the results are presented with the proper discussion. In the end, some concluding remarks are presented along with recommendations for future research