|
The development of applications for mobile autonomous systems to
carry out a determined task requires as basic assumption the
knowledge of the current position of the mobile system. To fulfill
this necessity, a lot of different methods have been proposed, some
more accurate than others. Therefore, when there are methods to
calculate a position estimation under different conditions and using
different hardware, the key is to find an estimation as accurate as
possible. Using a position estimation coming from one of this
methods, it is possible to apply an Extended Kalman filtering to
make the estimation better. This position introduced in the EKF does
not have to be so accurate and could depend on the conditions and
the available hardware. Moreover, if a group of robots which are
able to communicate among them are sharing the same environment, it
is possible to improve the accuracy of the position estimation using
data coming from the other robots and the relative distances to
these robots measured by some sensor. This is the basic principle of
the method developed in this master thesis.
|