Dissertation/Thesis Abstract

Stream function path planning and control for unmanned ground vehicles
by Daily, Robert L., Ph.D., Auburn University, 2008, 188; 3333176
Abstract (Summary)

This research develops a harmonic potential field based path planner and controller. The first component of the dissertation compares several analytical and numerical potential field generation methods. To overcome limitations present in the existing methods, two new methods are developed to meet the specific requirements of this research (generic shape obstacles and an explicit stream function). In particular, an analytic method to combine circles and create generic shaped obstacles is presented. Additionally, a numeric technique to directly calculate a potential field stream function is developed.

The second part of this research extends the traditional potential field controller to track a desired streamline as well as the potential field gradient. In addition to tracking a streamline, the reference path is also modified to maximize safety as the vehicle is driving. In particular, a separate harmonic potential field is created for the desired speed. This speed is low near obstacles and high in the open field. The lateral acceleration of the vehicle is also limited by reducing the steer angle and desired speed whenever an acceleration threshold is crossed. Finally, to create a buffer between the vehicle and obstacles, whenever the vehicle is close to an obstacle the desired streamline is shifted away from the obstacles. Together these three real-time modifications to the controller keep the vehicle safely away from obstacles and within its handling limitations.

Indexing (document details)
Advisor: Bevly, David M.
Commitee:
School: Auburn University
School Location: United States -- Alabama
Source: DAI-B 69/10, Dissertation Abstracts International
Source Type: DISSERTATION
Subjects: Automotive engineering, Mechanical engineering, Robotics
Keywords: Path planning, Stream function, Unmanned ground vehicles, Vehicle control
Publication Number: 3333176
ISBN: 9780549857402
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy
ProQuest