Towards aerial robotic workers : design, control and guidance for unmanned aerial manipulators

• Main Author: Imanberdiyev, Nursultan
• Other Authors: Pham Quang Cuong
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Engineering::Mechanical engineering::Robots
• Online Access: https://hdl.handle.net/10356/144236
• Institution: Nanyang Technological University

This thesis focuses on a novel aerial robot system composed of an unmanned aerial vehicle (UAV) and one or more robotic arms, which is employed for aerial manipulation tasks. In particular, this research tries to transform UAVs from a predominantly data gathering platform into a flying robot, frequently called as unmanned aerial manipulators (UAMs), that can physically interact with the object or surrounding environment. This is challenging due to the existing design limitations imposed by the attachment of robotic arms to floating platforms. Furthermore, adding on to the cross-coupled dynamics of the UAV platform and arms, the presence of strong nonlinearities, external disturbances (e.g., wind), and parameter variations further complicates the deployment of UAMs in day-to-day life. Moreover, having such redundant aerial robots, especially when employing two manipulators, imposes additional challenges to successfully perform UAM missions. Motivated by these challenges, the main contributions of this thesis are threefold. Firstly, we propose a novel design of the dual-arm manipulator specifically developed for UAM missions. The dual-arm system is designed such that it can be attached to different multirotors with minimum design modifications. In addition, the proposed dual-arm design employs prismatic joints to introduce three distinctive features: 1) ability of each arm to dynamically adjust its COG for better flight performance; 2) extended workspace and reach of the arms for enhancing operational capability and improving safety during UAM missions; 3) fully independent control of each arm for performing different tasks simultaneously. Secondly, we present a learning-based intelligent control approach, the fusion of artificial neural networks and type-2 fuzzy logic controllers, namely type-2 fuzzy neural networks, for the control of UAMs under time-varying working conditions. The proposed control strategy eliminates the need for precise tuning of conventional controllers, and it can compensate for the internal and external disturbances caused by the motion of the arms and unforeseen environmental changes, opening the door for the widespread use of UAMs in daily life. Thirdly, we propose two different trajectory generation with redundancy resolution strategies for UAMs. The first strategy is based on a weighted damped least-squares method, while the second approach employs a nonlinear model predictive control-based technique. The proposed approaches are capable to exploit the system redundancy by modifying the UAM configuration and simultaneously generate the feasible trajectories during the execution of the assigned tasks while enforcing the system constraints. The contributions mentioned above are evaluated throughout extensive simulation studies and experimental flight tests with real UAM platforms. Overall, these contributions aim for extending the autonomous functionalities and operational capabilities of UAMs in the transportation and assembly type of scenarios, with the ambition to assist lessening the current gaps that hinder the widespread deployment of aerial robotic workers in real-world applications.