Igs Stations' High-precision Solution Based On The Stochastic Model And Its Time Series Analysis

High precision IGS stations baseline vector calculation results can provide basic data for forecast of earthquake, geodynamic research, geophysics and space geodesy, but the signal was influenced in the process of transmission by many error sources, the baseline vector solution process can not form model, the solution precision can not satisfy our demand, so we joined three stochastic models in the resolving process to improve solution precision, the stochastic models which selected are:(1) the model based on the satellite altitude angle, (2)the model based on the signal to noise ratio, (3)the model based on the least square residual, then use the wavelet analysis method of time series, eliminate the gross error, analyze baseline vectors'variation law with time, thus we can get some useful conclusions. This paper selects four domestic IGS stations about BJFS (Beijing fangshan),SHAO (Shanghai sheshan),KUNM (kunming) and WUHN (wuhan), whose positions are relatively balanced, and quality of observation data is steady. The data which to be selected is the observation data of the four IGS stations of 2008. The main content of this paper:firstly introduces the origin of the research and the current research at home and abroad; then describes the basic principle of GPS positioning and the mathematical model of GPS precision positioning; then introduces the GAMIT software, and join the three kinds of random model in the solution, use the GAMIT software into the original observation data processing, finally using wavelet analysis method of time series to analyze the solution, from the results, we can see that this three kinds of random model can improve the calculation precision to different degree, the result of the model based on the least square residual is more closer to the ture value, its calculation is more reliable and can eliminate various errors much better. Thus, we found that by joining the stochastic model into the data processing can effectively improve the baseline vector solution precision, this also provides basis for modern GPS data processing.