| dc.description.abstract | Recent advancements in autonomous systems have significantly impacted both academia and industry, opening new research avenues. One of them is the collaboration of multiple systems to achieve a common goal, which is known as a cooperative system. In the lack of human intelligence, decision making and perception capabilities, uncrewed autonomous systems could mutually benefit from each other’s capabilities when they are deployed and utilized together. This research tackles the collaboration of group of uncrewed aerial systems (UASs) where the constraint on individual vehicles requires a varying level of coordination and cooperation. Such cooperation can be in the form of a physical support, where the task demands beyond the physical capabilities of a single system, and in the form of an intelligence level support, where a better perception, processing, or decision-making capability is needed in general. The objective of this study is the development and integration of cooperative guidance and control algorithms for a selected set of UASs and constrained mission scenarios which include the cooperative aerial payload manipulation task via multi-rotors with suspension cables and the cooperative formation task utilizing a team of airship and multi-rotors. Additionally, this research aims to integrate the developed algorithms for both individual and cooperative models in high fidelity simulations so that the effectiveness of multi-agent collaboration can be studied over realistic flight tasks.
The first part of the research focuses on the modeling and simulation of individual aerial systems. Systems considered in this research for the case studies include multi-rotors with a flexible-cable suspended payload and an airship. In this part, the mathematical models of these systems are derived by employing Euler-Lagrangian and Newton-Euler methods, respectively. The dynamics of flexible cable model are analyzed and compared with analytical catenary solutions. Furthermore, to improve the simulation accuracy, a momentum- and geometrical structure-preserving variational integrator is implemented for multi-rotors with flexible-cable suspended payload systems.
In the second part, guidance and control laws are designed for each individual system to provide attitude stabilization and trajectory tracking. Initially, a game-theoretic approach based on a linearized system model is investigated for attenuating the swing of the suspended payload. This approach considers various state feedback scenarios for the multi-rotor with a slung load system. Building upon the insights gained from these linear analyses, a catenary shape-informed geometric control approach is developed for the attitude and trajectory tracking control of this system. For the airship, both linear and nonlinear control methods are developed. These include a gain-scheduling based linear quadratic control and a nonlinear dynamic inversion (NDI) method, respectively. Both approaches are then compared against each other, focusing on their advantages and implementation ease.
Finally, cooperative guidance and control laws are developed for realistic scenarios with constrained mission objectives, requiring either physical or intelligence-level cooperation among a group of UASs. Drawing on the catenary analysis of flexible cables, a cooperative control scenario is constructed. This scenario demonstrates cooperation between vehicles for aerial manipulation of a suspended rigid payload using multi-rotors, where constraints stem from the payload capacity of a single vehicle and the physical connection of vehicles via flexible cables. Secondly, a leader-follower communication graph topology is employed in formation control scenarios involving a team of multi-rotors, highlighting the integration of an extended state observer (ESO) based total disturbance estimation model. This model significantly enhances the robustness of the system against external disturbances and unmodeled dynamics. Finally, we demonstrate the practical application of these studies in an illustrative scenario where cooperative formation support via UASs is needed in a search-and-rescue mission. In this scenario, we also utilize an airship to transport and deploy the multi-rotors to the mission destination where formation tasks are carried out adapting to various formation shapes and graph topologies. This scenario demands both physical and information level collaboration for enhanced area coverage, improved perception, and situational awareness. The constraints here arise from the physical limitations of individual vehicles (such as size, endurance, payload capacity, and operating environment) and information-level constraints (like processing power, sensing, and communication capabilities). This scenario forms a baseline that has practical applications in real life. | |
