Target search methods for space situational awareness

This work studies methods to detect target in an orbit around the Earth using a space based sensor. Searching for a target among a large set of candidate orbits is a difficult and time consuming problem. Considering orbital dynamics, sensor uncertainties and the initial size of candidate location distribution, it is desirable to develop efficient search techniques. In this work, information-theoretic methods for searching a target in a large probability distribution using a space based sensor is considered. One intuitive approach is to steer the sensor towards regions of high probability density. Alternatively, information-theoretic methods steer the sensor based on metrics of the information gain in the posterior probability distribution. Through simulation, it is shown that information-theoretic search methods produce greater knowledge about probability distribution of the target's orbit. We also present methods to lower the computing expense imposed on the computer on-board a space based sensor. The issue is addressed using data clustering technique called K-means clustering. It is shown that errors resulting from searching the target after clustering is much lower compared to errors resulting from searching targets at the locations of higher probability.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Aerospace Engineering

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2018

Author dc:creator

• Patel, Mihir Jagdishbhai

Contributors dc:contributor

• Ho, Koki

• Space situational awareness
• tracking
• target search

Statement dc:rights

• Copyright 2018 Mihir Patel

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/101218

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/101218