|
• Principal subjects/occupational
skills covered |
|
My degree thesis was entitled “Development and numerical
check of INS-stellar integration algorithms based on
Particle Filter”. It was about the study and the
simulation of an autonomous sensor, for determining a
long-term accurate satellite attitude estimation, based
on the integration between three inertial sensors and a
stellar one. Usually a that kind of integration it’s
realized by a Kalman Filter methodology, an optimal
Bayesian estimator suitable only for a restricted class
of linear Gaussian problems. In this case, because of
its suitability for nonlinear/non Gaussian processes,
it’s been introduced a Particle Filters (PF)
methodology, a suboptimal nonlinear filters class also
used for recursive Bayesian state estimation (especially
for target tracking application). It means that, in
order to make inferences about the dynamic system I was
analyzing, it’s been possible to use a nonlinear model,
describing the evolution of the state (attitude angles
in term of Eulero angles) with the time in a closer
manner to the real dynamic than a linear one does. First
the PF algorithm, that best-fitted the problem, was been
chosen and implemented (the filter parameters where
chosen trying to do a trade-off between accuracy and
computational effort). Second, the average accuracies
reached by the filter, varying the coupling between
inertial sensors (some models of sensors for each
Optics, MEMS and vibrating gyros were considered) and
different models of star trackers, were evaluated by
Monte Carlo simulations. Third, the PF was tested in
presence of a little perturbation and in a time-limited
absence of star tracker measurement. Fourth, filter
convergence time was calculated; then the Cramer-Rao
bounds were also evaluated to confirm the good
performance of the filter. Finally, the so-obtained
results were compared with those obtained utilizing the
more common and tidying up (but more restrictive) Kalman
Filter. The main related topics dealt with were been
aerospace systems, orbital mechanics, attitude
determination and control and statistics.
As regards my education in general, it was mainly
focused on both systems ( such as Systems Theory, Remote
Sensing, Aerospace Servosystem and Aerospace Systems)
and fluid dynamics subjects (Aerodynamic, Hypersonic
Aerodynamic, Experimental Aerodynamic, Thermo Fluid
Dynamics and Numerical Fluid Dynamics). |
