Seminarium i matematisk statistik
onsdagen den 9:e december 13.15 i MH227

Ola Björsne presenterar sitt examensarbete

Titel: Integration of Navigation Data

Abstract:

This Master's thesis presents investigations of the
possibility to improve the navigational data of submarines.
The improvements searched for are based on an extended Kalman
filter that is cascaded with the Inertial Navigation System
and includes the Dawid W. Taylor standard submarine equations
of motion as a dynamic model of the submarine. The extended
Kalman filter cofilters the measurements from the various
navigational subsystems into estimates of the position and
the states needed to control the submarine.

The measurement system consists of an Inertial Navigation
System, an Electromagnetic log, a Doppler log, a Depth gauge,
a Gyro compass and a GPS. Since a submarine has to surface
to use the GPS, the GPS is used as seldom as possible. The
filter is tested in simulated scenarios where the Doppler
log is used continously as well as in scenarios where it is
used only for short periods of time.

The report concludes that the configuration is a possible way
to improve the position estimate relative to the Inertial
Navigation System when Doppler log data is available. However,
when no Doppler log measurements are made, the low observability
of the dynamic model in the filter makes it difficult to
estimate changes in sea currents and the error in the position
estimate is in the same order of magnitude as for the Inertial
Navigation System.