Logo image
Robotic navigation for ductwork and close quarters
Thesis   Open access

Robotic navigation for ductwork and close quarters

Kristopher Krasnosky
Master of Science (M.S.), Drexel University
Mar 2014
DOI:
https://doi.org/10.17918/etd-4476
pdf
Krasnosky_Kristopher_20149.18 MBDownloadView

Abstract

Robotics--Plumbing Mobile robots--Automatic control Mechanical Engineering
This project aims to to study robotic navigation and performance in ductwork structures for a vehicle designed to map radiation levels as a function of position. A feasibility study was completed, examining the performance of a simple wheeled vehicle in ductwork. During this phase, various range finding, odometry and radiation detecting sensors were tested. Optical odometry combined with simultaneous localization and mapping (SLAM) proved particularly effective in this phase. A navigation study focused on navigation algorithms that could utilize SLAM and odometry data to navigate complex situations. Three main navigation techniques were studied: search tree navigation, Voronoi navigation, and potential field navigation. Basic potential field navigation was then extended to allow it to overcome local minimum in the potential function by using minimum free, harmonic functions as potential functions.

Metrics

37 File views/ downloads
12 Record Views

Details

Logo image