The Key Technology Research Of Precision GPS Positioning System-based Composite Differential Technique

Global Positioning System is simply referred to GPS, which has been widely used inpeople’s daily product and life. With the establishment and improvement of positioningsystems in various countries, it is now the era of co-existence of multiple constellations.GPS positioning system developed by the United States and another Global NavigationSatellite System, referred to GLONASS, developed by Russia are both more reliable andmore stable global positioning galaxy. The combination of the both system will have agood effect to improve the reliability, availability, and positioning accuracy. In this paper,the main research is GPS/GLONASS complex differential positioning. Using the relativedifference positioning, GPS/GLONASS complex relative differential positioning model isestablished. After data preprocessing, observation equation is solved. Thus, the rapid andprecision positioning is achieved. The accomplished specific content of the paper arefollows.Various error sources in positioning were analyzed. A mathematical model of theGPS/GLONASS complex differential positioning was established, which was based on theprinciple of GPS differential positioning and the relative differential positioningtechnology. Moreover, for the specific data formats of satellite positioning, RINEX andSP3, designed MATLAB-based reading program, and stored the related data in structurematrix, so that it was convenient for data processing.The precision ephemeris of GPS and satellite positions of GLONASS wererespectively extrapolated and fitted using the sliding type Lagrange polynomialinterpolation method and the fourth-order Runge-Kutta method to get the coordinates ofobservation satellites. Then the satellite elevation combined with the weighted geometricdilution precision (GDOP) to select the minimum galaxy for positioning. Relative to theminimum GDOP method, this method made the amount of calculation reduce87.23%.Proposed adaptive attenuation factor Kalman filter algorithm (AAFKF) for phasesmoothed pseudo range. Estimated the unknown pseudo-range noise variance by the noisestatistics estimator based on the maximum a posteriori, so improved the error due to the imprecise model. The algorithm was verified using measured data, and calculated thepseudo range poor of neighboring epochs. The results show that the algorithm has a goodeffect on smoothing pseudo range.On the basis of the least squares method, derived the iterative weighted least squaresmethod for solving the composite difference observation equations. It was a posterioriestimate variance components method which was used to determine the weighting matrixof GPS observations and GLONASS observations with different precision. Thendownloaded measured data, conducted simulation experiments about non-differential anddifferential positioning by GPS, GPS/GLONASS, weighted GPS/GLONASS. Finally, thedata from the experiments was analyzed. The results showed that the mean square error ofthree-dimensional coordinates getting from the weighted GPS/GLONASS complexdifferential positioning was less than10cm, and the solver speed of single epoch wasabout11ms.