Date of Graduation

8-2025

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Engineering (PhD)

Degree Level

Graduate

Department

Mechanical Engineering

Advisor/Mentor

Wejinya, Uche

Committee Member

Jensen, David

Second Committee Member

McCann, Roy

Third Committee Member

Shou, Wan

Keywords

robotics

Abstract

Two-wheeled self-balancing robots have garnered substantial attention within the realms of research and innovation in academia and industrial settings. In particular, advances in control algorithms, machine learning, reinforcement learning, and sensor technologies have played a pivotal role in their development. Although primarily recognized for their utility in personal transportation in the commercial sector, these robots possess potential applications across various domains, including search and rescue, healthcare, material handling, logistics, etc., due to their active stabilization and exceptional maneuvering capabilities. A particular area of interest lies in the creation of modular self-balancing robots designed for seamless reconfiguration and enhanced human interaction. Hence, this dissertation aims to conceive a modular self-balancing robot that employs advanced control strategies to effectively tackle the challenges associated with parametric uncertainty, ensuring that it maintains equilibrium and navigates adeptly even in dynamically changing environments. The proposed modular robot design incorporates key components such as microcontrollers, an inertial measurement unit (IMU), motors, motor drivers, and interchangeable components, such as storage units, sensory modules, and actuation modules, allowing for swift reconfiguration tailored to specific applications. This modular approach amplifies the adaptability and versatility of the robot, enabling it to excel in various scenarios. This dissertation places emphasis on the development and implementation of advanced control strategies, which encompass adaptive control and reinforcement learning, to effectively manage uncertainties that stem from variations in parameters such as mass, friction, and center of gravity. The development of a modular self-balancing robot attempts to highlight the significance of modularity and adaptability in the design of robotic systems that must operate in dynamic and uncertain conditions.

Download

Share

COinS