• Batch Least-Squares Algorithm
• Kalman Filter
• Orbit Determination

In this thesis different orbit determination methods are implemented and their performance are evaluated and compared. Specifically, batch least-squares, linearized Kalman filter and extended Kalman filter are considered. The aim is to evaluate the performances of the three orbit determination methods when the measurements provided by a ground tracking station are used.
As practical application, the problem of orbit determination is applied to the Client spacecraft of the DEutsche Orbitale Servicing Mission (DEOS); the two possible configurations of the orbit of the satellite (altitude of 400 km or 600 km) have been taken into account. The tracking station performing the measurements for the DEOS mission is the Forschungsgesellschaft für angewandte Naturwissenschaften (FGAN) located in Wachtberg, near Bonn.
Since measurements from the FGAN station are not available, simulated measurements have been realized. To this purpose an orbit propagator has been used to obtain the orbit of the Client. From this orbit simulated measurements have been derived considering the measurement error introduced by the FGAN station; complete and precise data for this error are however not available.
The simulated measurements have been used as input to the three implemented orbit determination methods. The results provided have been analysed considering the sensitivity to the accuracy of the reference orbit, to the number of measurements included in the process and to the error introduced by the measuring device. A comparison of the three methods has indicated a good agreement of results provided by batch least-squares and Kalman filters.