|
Global Positioning System (GPS) has become the leader of
modern electronic navigation systems due to its properties of
high accuracy and multi-purpose.However, the GPS is also a
physical system, so the positioning accuracy willdecrease
because of the disturbance error source. From relevant
literature, Kalman Filter, which is a multi-input/multi-
outputsystem, can be successfully applied to GPS. Kalman Filter
can not only computethe position and clock statesof the receiver
simultaneously, but also its recursive form can easily be
realized to avoid the increase of computation load.Besides, we
can make use of the "off-line analysis process" to evaluate the
positioning accuracy and the process models of Kalman Filter
before the posi- tioning estimation starts. When we estimate
the states of a dynamic receiver, it is often to describe the
dynamics of the receiver with a process model called "Position-
Velocity model"or "PV model". But it is inadquate to describe
the high dynamics receiver with PV model because velocity
changes rapidly in a short time interval. The purposeof this
thesis is to describe the high dynamics receiver with anoyher
processmodel called "Position-Velocity-Acceleration model" or
"PVA model" used in Kalman Filter for improveing the positioning
accuracy. Prediction results from off-line error analysis show
that the ability of PVA model to improve the GDOPChimney due to
bad observation conditions is better than PV model. Simulation
results also show that PVA model can better describe the
dynamics of a high dynamic receiver, as a result, wether the
observation conditions is, the posi-tioning accuracy of PVA
model is always better than PV model.
|