Sensor Tasking for Satellite Tracking Utilizing Observability Measures
View/ Open
Date
2018-12-06Author
McDonald, Mitchel T.
0000-0002-8836-2449
Metadata
Show full item recordAbstract
In this thesis, novel observability-measure based sensor tasking methods are studied for satellite tracking applications. The tasking is performed by first computing the Hellinger Distance between ground/space based sensors and space objects, and then the measure is utilized for selecting the sensors that maximize observability. Several other measures such as the Fisher Information Gain, Largest Lyapunov Exponent, and Shannon Information Gain have also been utilized and the performance of these measures are computed against each other. The object's state estimates are obtained using nonlinear estimation techniques. The Extended Kalman Filter, Unscented Kalman Filter, and Bootstrap Particle Filter are compared within in this framework. Representative numerical simulations are performed to evaluate the efficacy of the new tasking approach. The proposed tasking approach is also compared with some baseline approaches.