Access Type
Open Access Dissertation
Date of Award
January 2024
Degree Type
Dissertation
Degree Name
Ph.D.
Department
Mechanical Engineering
First Advisor
Azad Ghaffari
Abstract
This dissertation explores, investigates, develops and experimentally validates algorithms for safe and robust motion control of ground robots and vehicles. The emphasis is on design of controllers for ground autonomy with safety either embedded by design or enforced by appropriately wrapping a safety net around an existing nominal controller.
Ground robots have been considered for their importance in compensating for labour or personnel shortage for material handling in the manufacturing, logistics and healthcare sectors. On the other hand, ground vehicles have been considered for the existing potential of making driver assistance systems safer and more intelligent, particularly in at-the-limit driving scenarios. For ground robots, safety is considered in the context of obstacle-avoidance. For ground vehicles, safety and robustness is considered in the context of retaining tire-road traction in safety-necessitated aggressive maneuvers, despite the presence of modeling errors.
Throughout this dissertation, the control architectures and schemes developed are minim\-ally invasive, that is, the control effort of an existing nominal controller is modified only if safety is expected to be compromised. This has promoted adaptability, modularity and compactness in the design of the control architectures.
Where considered appropriate, the control design in this dissertation is accompanied by theory and analyses for provable safety and stability. This enables their implementation on real-world robots in the near-term. For the long-term, the theory and analyses provided lay the foundation for developing data-driven controllers that are embedded, by design, with stability and safety guarantees.
The main contributions of this dissertation are,
1. provision of a control architecture that decouples kinematic and dynamic control of ground robots in a loop-within-a-loop fashion for trajectory-tracking and obstacle-avoidance. At the kinematic level, a self-adaptive barrier function-based safe steering envelope provides safe commands for translational velocity and heading. At the dynamic level, the innermost wheel velocity control loop comprises a compensator for actuator-level time-delays that allows timely steering for obstacle avoidance,
2. formulation of a minimally invasive auxiliary control that avoids undesirable local equilibria in control Lyapunov function and control barrier function based quadratic programs designed for trajectory-tracking and obstacle avoidance in ground robots. The auxiliary control is accompanied with specific QP constraints designed based on the theoretical investigations that provide the mathematical conditions for occurrence of the undesirable equilibria, 3. development of an accurate and robust input-to-state stable nonlinear observer for ground vehicles that can estimate, online and in real-time, errors in modeling lateral and longitudinal vehicle dynamics, and translate these errors into improved estimates of ground-truth tire-road forces. Experimental validation on a one-tenth scale and full-scale cars opens up opportunities to improve vehicle-tire modeling and robustness of vehicle control to ultimately improve vehicle safety,
4. construction of an observer robustified control barrier function filter that wraps around a nominal vehicle controller to retain traction and motion control in a vehicle driven at the limits of handling, despite the presence of modeling errors in the nominal controller. Experimental validation on a full-scale vehicle showed improvement in vehicle safety by preventing loss of tire-road traction and consequent vehicle understeer and oversteer.
Recommended Citation
Desai, Manavendra, "Algorithms For Safe And Robust Motion Control Of Autonomous Ground Robots And Vehicles" (2024). Wayne State University Dissertations. 4036.
https://digitalcommons.wayne.edu/oa_dissertations/4036
