Theses and Dissertations

ORCID

https://orcid.org/0000-0001-9434-3570

Advisor

Luo, Chaomin

Committee Member

Follett, Randolph F.

Committee Member

Choi, Seungdeog

Committee Member

Karimi-Ghartemani, Masoud

Date of Degree

8-13-2024

Original embargo terms

Visible MSU Only 2 Years

Document Type

Dissertation - Campus Access Only

Major

Electrical & Computer Engineering

Degree Name

Doctor of Philosophy (Ph.D.)

College

James Worth Bagley College of Engineering

Department

Department of Electrical and Computer Engineering

Abstract

In recent years, significant progress has been made in autonomous robotics and the electrification of transportation, highlighting the growing importance of automation in daily life. Ensuring the safety and sustainability of automated systems necessitates the integration of intelligent algorithms capable of making astute decisions in uncertain circumstances. Autonomous robots possess considerable potential for efficiently performing intricate tasks, but this potential can only be unlocked through intelligent algorithms. Moreover, enhancing the energy efficiency of transportation systems yields extensive benefits for the environment, economy, and society at large. Addressing the urgent challenges of climate change and resource depletion necessitates prioritizing energy efficiency in transportation to construct a more resilient and equitable future. This research delves into the development of bio-inspired neural dynamics, nature-inspired swarm intelligence, fuzzy logic, heuristic algorithms, and optimization techniques for optimal control and navigation of electrified and unmanned ground vehicles. Drawing inspiration from biological systems, this research aims to enhance the performance of robots in dynamic and unstructured environments. The approach encompasses a hybrid bio-inspired method, leveraging the mathematical model of a biological neural system's membrane to facilitate smooth trajectory tracking and bounded velocities for a differential drive robot. Additionally, integration of a Leader-Slime Mold Algorithm (L-SMA) for global path optimization and a modified velocity obstacle (MVO) for local motion planning is pursued. A heuristic algorithm is also devised to enhance decision-making in uncertain and dynamic environments by coordinating actions among the L-SMA path planner, the MVO local motion planner, and the enhanced bio-inspired tracking controller. Furthermore, a real-time optimal predictive controller is proposed to address the energy management challenges of electrified vehicles while improving driveability and comfort. This predictive controller employs a linear parameter-varying model of an electrified vehicle, a custom-designed adaptive cost function, and fuzzy logic to adapt a subset of cost function weights. The integration of fuzzy logic and the adaptive predictive controller yields a convex optimization problem solved in real-time using an active-set solver. To further enhance the energy efficiency of the electrified vehicle, a particle swarm optimization enhanced model predictive controller is suggested as an adaptive cruise controller with superior energy efficiency and safety in vehicle-following scenarios. Through these integrated approaches, the aim is to advance the capabilities of autonomous robotics and electrified transportation systems, thereby contributing to safer, more efficient, and sustainable mobility solutions.

Download

Share

COinS