The Improved Random Model In Combined GPS/BDS Positioning Precision

The United States、Russia、Europe have established and developed the GPS, GLONASS, Galileo satellite navigation and positioning system. Not far behind them, as a rising star for the development of navigation system, China independently established the beidou satellite navigation system with its intellectual property rights. At this point, the development of a mixture of multiple satellite navigation systems has been initially formed, compatible GNSS in Multimode、its interoperate and Multimode GNSS data fusion will become one of the main development trend in GNSS navigation field. According to the fact that how to determine the weight of observation data from different system when joint positioning with the multi-system of GNSS, this paper analyzed a prior stochastic model including the geometric, satellite altitude Angle and PDOP value analysis stochastic model etc; Posterior variance component estimation for the influence of joint positioning accuracy. Refer to the inside and outside accuracy of the three-dimensional position as the single point positioning performance indicators, there are several research contents as follows:Firstly, campare the point positioning quality of GPS and BDS in experimental areas, then campare the results of single point positioning between the two systems, get the basic conditions of GPS/BDS joint positioning, make a reliable analysis of the results, and provide reasonable explanation;Secondly, I have discussed how to detemine the weight matrix of the GPS/BDS combined system. By comparing the equal weight model with the elevation angle model and PDOP value analysis stochastic model and Helmert posterior estimation, pointing out that deterning the weight matrix of the GPS/BDS combined system based on Helmert posterrior estimation is the most propable.Thirdly, set the initial weight matrix by different priori stochastic models, get the single point positioning result of GPS/BDS, then combined the Helmert posterior estimation, conclud the stability and reliability of the ultimate single point positioning resulet combined the Helmert posterior estimation under the condition of different priori stochastic models.Fourthly, count the available satelites number of the single system and combined system, analyse the reason of different positioning performance between different systems.Numerical examples show that: the accuracy of single point positioning between GPS and BDS has the same level, beidou pseudorange single point positioning accuracy level is higher than 6 m, height accuracy is better than 10 m; Campared with the GPS system, the positioning accuracy of GPS/BDS combined system got a improvement. In a complex environment, GPS/BDS integrated system can still meet the number of satellite positioning, and keep a certain accuracy, because of the satellite distribution structure was optimized; Compared with using only one priori stochastic model on weighting, a priori and posteriori random model combination made by positioning results have been improved significantly, and the positioning results using the satellite angle model considering PDOP value combined with the posteriori variance component estimation method have the highest precision, resistance, and stability.The results of this article have a certain reference value in multiple system joint position and the improvement of positioning accuracy and reliability of beidou system after network.