| One of the disadvantages of an inertial navigation system (INS) is its errors will grow unbounded. The global positioning system (GPS) can be used to improve position estimation resulting from INS measurement. The Kalman filter method is used to fuse measurements from the systems (INS and GPS). When the channels available in the GPS receiver are restricted, the receiver has to choose the combination of satellites, which will give the lowest noises in the GPS measurement signal. In this thesis, the method to find the optimal combination of the satellites has been examined. To effect the change of satellite combination an error estimation has been investigated and presented. The comparison of the implementation of indirect half feedback and indirect feedforward Kalman filter configurations in an aircraft has been performed and analyzed. The GPS and INS measurement signals used in this experiment are real data, which was recorded on an aircraft during the flight. The experiment shows the relation between combination of the satellites and the estimated position and velocity of the aircraft. The improvements in the indirect half feedback configuration are also achieved. |
