Information based sensor control in a two-vortex flowfield

"This thesis is divided in two parts. First, after presenting the modeling of the system in the deterministic and stochastic case, the controllability of one, then two Unmanned Aerial Vehicles (UAVs) manoeuvring in a two-vortex flow field is proven in the nonlinear case. In the second part, we develop a complete algorithm for optimally controlled sensor platform in a vortex flowfield. The question raised is: ""How to control a UAV network evolving in a set of two vortices to collect the best data from these vortices?"". The control part is based on information theory where the cost function is built using the the Kullback Leibler measure. The sensor platform is evolving in a vortex environment which is not controlled, yet has random initial conditions. Also, it is assumed that the UAVs are observed and that their positions are completely known. Thus we consider their dynamics as the only information available on the vortices. The sensor platform is in charge of finding the location that will contain the best information before actually getting the measurements from that place. We collect exclusively useful data by tracking the position that is expected to contain the best information using a specific information metric."

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

2013

Author dc:creator

• Mamlouk, Mahmoud

Contributors dc:contributor

• Namachchivaya, N. Sri

• Unmanned Aerial Vehicle (UAV)
• extended Kalman filter
• data assimiliation
• Kullback Leibler measure

Statement dc:rights

• Copyright 2013 Mahmoud Mamlouk

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/45494

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/45494