Research On Nonlinear Algorithm For Improving GPS Positioning Precision

Global Positioning System (GPS) is the only perfect kind of new-generation precise satellite navigation system. GPS has been widely concerned on military and the civilian since it was used, which has been developed rapidly, for its characteristics of providing all-global, continuous, real-time and high accuracy 3-D positioning and velocity as well as time. Currently, it is widely applied in many fields, such as aerospace, aviation, maritime, vehicle navigation, geographic positioning, surveying and so on.With the development of navigation and positioning, in recent years, people have higher demands for the GPS positioning precision. Now, the highest accuracy of stand-alone GPS receiver is about 10 to 15 meters, and when GPS receiver was in the urban canyon, the number of satellite signals which was received may be less than 4, so the positioning was impossible.The positioning accuracy of stand-alone GPS receiver is low since there are many error sources during the signal transferring from GPS satellite to GPS receiver. The focus of this thesis was on measurement model and positioning algorithm. The GPS positioning accuracy was improved by established completed error rectification models and used high precision algorithm.The work of this thesis was as follow:1) The rectification models of ionosphere and troposphere were established to improve the positioning accuracy of GPS receiver.2) As the velocity of vehicle was low and the position changed was small in the urban canyon, hypothesis was address that the position of GPS receiver in direction of height was fixed. Then height equation was added to measurement equations so 2-D GPS positioning was implemented.3) Using Unscented Kalman Filter (UKF) as GPS positioning estimation algorithm instead of Iterative Least Square (ILS) and Extended Kalman Filter (EKF).4) The Spherical Unscented Kalman Filter (SUKF) was adopted to improve the speed of GPS positioning estimate calculation. This algorithm can be used for outputting real-time position. 5) In order to improve the positioning accuracy further, smooth method was adopted. Take consideration to output real-time position, iterative Kalman Filter and Kalman Smoother algorithm was addressed.The research of this thesis can be used in GPS software receiver to output high accuracy real-time positioning results. It also can be used to improve the positioning estimate method in high-quality GPS receivers.