Graduation Year

2020

Document Type

Thesis

Degree

M.S.E.E.

Degree Name

MS in Electrical Engineering (M.S.E.E.)

Degree Granting Department

Electrical Engineering

Major Professor

Wilfrido A. Moreno, Ph.D.

Committee Member

Ismail Uysal, Ph.D.

Committee Member

Chung Seop Jeong, Ph.D.

Keywords

Convolutional Neural Networks (CNN), Indoor Positioning, Mobile Robot, Q-Learning, Skid-Steering

Abstract

Autonomous robots are employed in numerous areas. In this thesis, it is proposed to design and build a self-controlled wheeled vehicle to deliver food.

As there are many applications of an autonomous agent for indoor and outdoor environments, this study is conducted on indoor settings whereas all the requirements and design processes are achieved for both operational boundaries.

The fundamental approach is to design and implement a Wheeled Mobile Robot (WMR), and to test skid-steering performance on proposed trajectories using a System Engineering approach. From this point of view, system requirements in mechanical, electrical, and software are evaluated, and overall system is divided into subblocks which are motor processor, image processor, and central processor.

One of main concerns is indoor and outdoor positioning. While outdoor tasks are widely solved in terms of the Global Positioning System (GPS) technology, indoor navigation appears with challenges. Hence, it is aimed to acquire a deeper understanding in mobile robot indoor localization through Deep Neural Networks (DNN) and learning algorithms.

Download

Share

COinS