- 60 views
- 111 downloads
Nonlinear Control of Aerial Manipulators
-
- Author / Creator
- Jiang, Zifei
-
With advancements in the autonomy and capabilities of Unmanned Aerial Vehicles (UAVs), their potential for complex operations such as infrastructure inspection, maintenance, and transportation has increased. This thesis delves into innovative methods to enhance UAV autonomy, focusing on the motion control and force control of specialized UAV configurations. The study primarily revolves around the dynamics and control of slung load systems, the implementation of quasi-static feedback linearizing control, immersion and invariance adaptive control for fully-actuated aerial manipulation, and the application of reinforcement learning for improved UAV performance.
One key area of this research is the intricate dynamics associated with slung load systems in UAVs, which are integral for complex transportation and deployment operations. These systems demand precise control for managing the dynamics of their cargo. The thesis introduces a quasi-static feedback linearizing algorithm, universally applicable to differential flat systems. Based on this algorithm, quasi-static feedback controllers for a slung load system are developed, making the slung load system precisely linearized. The quasi-static feedback controller is adept at managing both the outer-loop and the full system dynamics of the slung load system. The controller's effectiveness, robustness, and accuracy in trajectory tracking are validated through extensive experiments in simulated and real-world scenarios, significantly enhancing UAVs' ability to handle heavy loads, a key aspect in diverse operational settings. Additionally, the research contributes to the development of a Maple-Matlab Software-In-The-Loop pipeline, streamlining the creation of nonlinear UAV controllers.
Another pivotal aspect of this thesis is the formulation and application of immersion and invariance adaptive control in fully-actuated aerial manipulators. Fully-actuated aerial manipulators, equipped with rigidly connected sticks, are becoming vital in tasks that necessitate direct environmental interaction, such as maintenance or disaster recovery. Integrating immersion and invariance adaptive control with these aerial manipulators addresses the complexities of interacting with unpredictable environments. The newly developed immersion and invariance adaptive hybrid force-motion controller demonstrates global asymptotically stable results, surpassing classical adaptive control theories by not requiring persistent excitation conditions or linear parameterization assumptions. This innovative control strategy ensures stable, robust, and efficient manipulation, unlocking new possibilities for sophisticated and autonomous aerial manipulators.
Furthermore, the thesis explores the application of reinforcement learning to enhance UAV autonomy. This segment focuses on developing learning algorithms that enable UAVs to adapt to diverse environments and tasks autonomously. The use of reinforcement learning allows UAVs to learn from their flight data, leading to continuous improvement in performance and adaptability. The effectiveness of these learning strategies is demonstrated through various simulations, showcasing their potential in advancing UAV capabilities.
In summary, this thesis presents a comprehensive study on advancing UAV technology through the control of slung load systems, the application of quasi-static feedback linearizing and immersion and invariance adaptive control for aerial manipulation, and the innovative use of reinforcement learning. These contributions mark a significant step forward in the field of UAV technology, expanding their applications in complex and dynamic environments.
-
- Subjects / Keywords
-
- Graduation date
- Fall 2024
-
- Type of Item
- Thesis
-
- Degree
- Doctor of Philosophy
-
- License
- This thesis is made available by the University of Alberta Library with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.