Advanced

Information-Theoretic Approach for Path Planning of a Moving Platform with Bearings-only Sensor

Björström, Rickard (2004) In MSc Theses
Department of Automatic Control
Abstract
Using flying vehicles for reconnaissance and surveillance has always been interesting, especially in military applications. Unmanned aerial vehicles (UAVs) have been increasingly used in the last decades, but they have often been controlled by an operator on the ground. In an attempt towards higher level of autonomy, the UAV should be able to decide itself where to fly. This thesis examines a method for autonomous path planning based on the uncertainty of the target locations, a so called information-theoretic approach. A bearings-only sensor is attached to the UAV, such as a video or an infrared sensor, which makes observations of the relative angle to the object, reducing the uncertainty orthogonal to the observed target. The planned... (More)
Using flying vehicles for reconnaissance and surveillance has always been interesting, especially in military applications. Unmanned aerial vehicles (UAVs) have been increasingly used in the last decades, but they have often been controlled by an operator on the ground. In an attempt towards higher level of autonomy, the UAV should be able to decide itself where to fly. This thesis examines a method for autonomous path planning based on the uncertainty of the target locations, a so called information-theoretic approach. A bearings-only sensor is attached to the UAV, such as a video or an infrared sensor, which makes observations of the relative angle to the object, reducing the uncertainty orthogonal to the observed target. The planned path is the solution to an optimization problem, such that the uncertainty is minimized which is equal to maximizing the information in the information-theoretic approach. When identifying a target, there is a potential benefit to make observations from different views. If the path could be planned, the better are the observations of the target, and the image based identification will be more reliable to target appearance variations and more robust against decoys. (Less)
Please use this url to cite or link to this publication:
author
Björström, Rickard
supervisor
organization
year
type
H3 - Professional qualifications (4 Years - )
subject
publication/series
MSc Theses
report number
TFRT-5727
ISSN
0280-5316
language
English
id
8848020
date added to LUP
2016-03-18 19:03:47
date last changed
2016-03-18 19:03:47
@misc{8848020,
  abstract     = {Using flying vehicles for reconnaissance and surveillance has always been interesting, especially in military applications. Unmanned aerial vehicles (UAVs) have been increasingly used in the last decades, but they have often been controlled by an operator on the ground. In an attempt towards higher level of autonomy, the UAV should be able to decide itself where to fly. This thesis examines a method for autonomous path planning based on the uncertainty of the target locations, a so called information-theoretic approach. A bearings-only sensor is attached to the UAV, such as a video or an infrared sensor, which makes observations of the relative angle to the object, reducing the uncertainty orthogonal to the observed target. The planned path is the solution to an optimization problem, such that the uncertainty is minimized which is equal to maximizing the information in the information-theoretic approach. When identifying a target, there is a potential benefit to make observations from different views. If the path could be planned, the better are the observations of the target, and the image based identification will be more reliable to target appearance variations and more robust against decoys.},
  author       = {Björström, Rickard},
  issn         = {0280-5316},
  language     = {eng},
  note         = {Student Paper},
  series       = {MSc Theses},
  title        = {Information-Theoretic Approach for Path Planning of a Moving Platform with Bearings-only Sensor},
  year         = {2004},
}