Reseach On The Estimation Methods Of Real-Time Precise Satellite Clock Bias

With the development of geodesy and its related disciplines, the PPP has become the key problem in the field of current GPS positioning study, and precise satellite clock bias and real-time orbit products become the main factor of restricting the PPP real-time positioning technology. The cm level result of precise point positioning (PPP) can be provided by IGS precision products, and the real-time track date can be got by IGU. Research results indicates that the difference between IGU orbit and the precision of the IGS final orbit is small. However, because of GPS satellite clock with rubidium standard and cesium clock, they are vulnerable to the effects of the noise of the clock and the frequency shift. The change of satellite clock bias is very complex so that it is not easy through the model to make an accurate prediction, and eventually led to the difference of IGU clock accuracy is too low and cannot be applied to some relatively high precision in real time single point positioning. Therefore, the most critical issue in the study of the PPP is the real-time access of precise satellite clock error.The basic theories and methods of satellite clock offset estimation are systematically investigated in the thesis. The major corrections and related mathematic models in calcul-ating precise satellite clock offsets are discussed in detail. Through the c# programming language to realize the model of besed on difference observation combination equation of square root kalman filter. By comparing calculated clock difference and IGS precision clock difference, the absolute value is less than 1 ns, rarely exceeds 0.5 ns-0.5 ns. The precision evaluation value is mostly less than 0.5 ns by analysis the calculating results of different satellites for three consecutive days. This thesis also studies the influence of the clock difference calculating result by the different number of station and station distribution by using the date of America IGS stations. When the station number more than 28, the calculating results has been basically stable, mostly less than 0.4 ns. When the distance between the station more than 500 km, the correlation of satellite clock error and the troposphere parameters can be significantly reduced, much precision less than 0.4 ns.