Autonomous navigation via measurement of natural optical signals
Abstract
A method for using natural light sources for navigation and geolocation is described, and the general properties of a sensor capable of accurately measuring the locations of these light sources are explored. Natural light sources include the sun, moon, and stars, but local point sources of light (e.g. radio tower lights, airport lights) with known positions may also be used to enable accurate localized navigation. The navigation method described is highly susceptible to noise and uncertainties in certain regions, and is not suitable for general purpose navigation. However, the resolution is sufficient to enable a hybrid approach to work. This hybrid approach involves the computation of a rough estimate of the observer's absolute position on Earth followed by using local references for precise navigation, similar to how homing pigeons are known to navigate. Iterative and non-iterative solutions for the navigation equations are examined. Additionally, a novel solar sensor geometry is proposed which utilizes a hemispherical arrangement of pixels to determine the vector to the dominant light source in its environment. Two methods for calculating this vector are described: one based on transforming the problem from an optical one into a geometric one, and another using a least squares approach to minimize the error function for the sensor. These methods do not depend on the physical details of any particular sensor design. Both methods are derived under certain idealized assumptions, but methods for extracting an accurate sun vector are discussed which do not require such an ideal environment. Using the estimate of the sun vector and knowledge of the time, it is possible to determine the location of the sensor on the Earth. Applications include position measurement and navigation for UAVs, UGVs, and other mobile platforms.
Degree
M.S.M.E.
Advisors
Ariyur, Purdue University.
Subject Area
Mechanical engineering|Astronomy|Optics
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.