GNSS Positioning Solving And Accuracy Analysis Basing On Weighted Least Squares Algorithm

Global Navigation Satellite System(GNSS)has become one of the fastest growing information industries in the world.GNSS have played an important role in both civil and military aviation.It can provide positioning,navigation and timing services in a global,all-weather,continuous and real-time situation.However,in several areas,for example,the approach segment of aircraft,the accuracy of satellite positioning still can't satisfy their requirements,the improvement of the satellite positioning accuracy is still a research topic.The satellite positioning solving algorithm has a direct influence on the positioning result,so the study of satellite positioning solving algorithm help to fully play the role of GNSS.Basing on the conventional least squares positioning solving algorithm,this paper studies the weighted matrix of weighted least squares positioning solving algorithm,including following two parts:1.For single satellite system positioning,introducing the Weights Estimation Model basing on satellite elevation and user range accuracy,the Posterior Weights Model basing on the deviation of pseudo-range,the compositions of these two models and the implementation processes are given.Then through GPS and Beidou data from receiver to do the experiment,for the positioning results,using the average positioning error of each axis,Circular Error Probable and Sphere Error Probable these three parameters to do accuracy analysis to prove the validity of these weighed model.2.For combined satellite system positioning,firstly,unify the time and space system of different satellite system.According to Hermelt variance component estimation method,introducing the Posterior Variance Model to set the weighted matrix,then combining this model and Weights Estimation Model to design Fusion model,finally through the GPS and Beidou data to do the experiment,and using the Dilution of Precision,the average positioning error of each axis,Circular Error Probable and Sphere Error Probable these four parameters to do accuracy analysis for positioning results.The experimental results show that for single satellite system positioning,the positioning accuracy of Weights Estimation Model and Posterior Weights Model is better than conventional least squares positioning solving algorithm;for combined satellite system positioning,the positioning accuracy of combined satellite system is better than single satellite system,comparing with the least algorithm,the Posterior Variance Model can improve the positioning accuracy and the Fusion model can improve more the accuracy.