Dissertation
Coordinated control strategies for safe execution of multi-vehicle autonomous missions
University of Iowa
Doctor of Philosophy (PhD), University of Iowa
Autumn 2024
DOI: 10.25820/etd.007574
Abstract
Autonomous Vehicles (AVs) are becoming an integral part of our daily lives, playing crucial roles in both civil and military sectors. Cooperative fleets of AVs, in particular, offer several advantages in terms of robustness and efficiency. Despite significant advancements in this research area, a lack of safety guarantees still limits the employment of collaborative missions for many real-world applications, especially in complex environments.
To address this limitation, this thesis presents robust and efficient coordination and collision avoidance algorithms to enable a fleet of AVs to safely and efficiently navigate their environments. Two distinct speed adjustment solutions for coordinated control are proposed, accommodating different assumptions about the communication capabilities of the agents. These algorithms are designed to be applicable to heterogeneous groups of vehicles, enhancing flexibility, scalability, and cost-effectiveness. Additionally, this work addresses the practical challenges inherent in real-world applications, such as the dynamic limitations of vehicles, constrained communication capabilities in terms of range and bandwidth, and the issues posed by noisy or intermittent sensing. Addressing these issues is vital for the practical applicability and operational viability of the proposed solutions. It is shown that together with trajectory generation and target tracking algorithms, the control laws proposed in this thesis provide a cooperative control framework for safe execution of multi-vehicle autonomous missions.
Using tools from nonlinear control, Lyapunov theory and algebraic graph theory, we derive performance bounds, and define the convergence rate of the coordination dynamics as a function of the quality of service of the underlying communication network. Finally, the effectiveness of the introduced algorithms is validated through both simulations and experimental testing, confirming their efficiency and applicability to real-world scenarios.
Details
- Title: Subtitle
- Coordinated control strategies for safe execution of multi-vehicle autonomous missions
- Creators
- Camilla Tabasso
- Contributors
- Venanzio Cichella (Advisor)Isaac Kaminer (Committee Member)Rachel Vitali (Committee Member)Shaoping Xiao (Committee Member)
- Resource Type
- Dissertation
- Degree Awarded
- Doctor of Philosophy (PhD), University of Iowa
- Degree in
- Mechanical Engineering
- Date degree season
- Autumn 2024
- DOI
- 10.25820/etd.007574
- Publisher
- University of Iowa
- Number of pages
- xii, 125 pages
- Copyright
- Copyright 2024 Camilla Tabasso
- Language
- English
- Date submitted
- 12/02/2024
- Description illustrations
- illustrations, graphs
- Description bibliographic
- xii, 125 pages
- Public Abstract (ETD)
- In the last decade, Autonomous Vehicles (AVs) have played an increasingly bigger role in our daily life. Ranging from self-driving cars to search and rescue drones, AVs are becoming integral to many aspects of modern society. Cooperative autonomous fleets, in particular, offer immense potential to revolutionize tasks classified as ”Dirty, Dull, and Dangerous,” thereby enhancing safety, efficiency, and reducing operational costs. However, ensuring the success of missions involving multiple AVs presents a considerable challenge due to frequent interactions between vehicles and obstacles which could lead to collisions. The overarching objective of this work is to advance safety measures in collaborative autonomous missions and establish a robust framework for the seamless integration of AVs into our daily lives. In this thesis, we introduce cutting-edge algorithms for coordinated control, focusing on collision avoidance strategies. Central to this approach is the implementation of control laws that modify the vehicles’ speeds to prevent collisions, while trying to minimize deviations from the original mission plan. Diverging from existing solutions found in the literature, we design decision-making strategies for AVs focusing on their applicability to real-world scenarios. This is achieved by formulating realistic assumptions about the agent’s ability to perceive and analyze their environment and to exchange information promptly and dependably. Through a comprehensive evaluation that encompasses both theoretical analysis and experimental testing, we show that the algorithms presented in this thesis enable a fleet of AVs to safely and efficiently navigate their environment and mark a step toward a wider adoption of autonomous vehicles.
- Academic Unit
- Mechanical Engineering
- Record Identifier
- 9984774457002771
Metrics
4 Record Views