Research On Stochastic Model For Gps Precise Positioning

Zhicheng Lv,Zengjun Liu,Guozhu Zhang,Feixue Wang

2011-01-01

Abstract:With the fast development of GPS measure technology, the stochastic model has now being the critical ones in restricting GPS baseline result towards more precise. Taking into account the heteroscedasic, space correlation and time correlation property of GPS observation data, the inaccurate stochastic model may lead to the baseline result unprecise and unavailable. In this article, the stochastic modeling of GPS data processing is deeply discussed. The advantages and disadvantages of several GPS stochastic models such as equal weigh model, satellite elevation angle model, SNR model and posterior estimation model are compared. Finally, different stochastic models are compared and the result of the reliability of ambiguity resolution and the precision of positioning, which is evaluated by the ratio method and posterior estimation of variance-covariance method respectively, is list. The theoretical and numerical results show that with the MINQUE(Minimum Norm Quadratic Unbiased Estimation) method, the influence of the remaining systematic and pseudorandom noise on the positioning result could be effectively removed and the final positioning equality were improved.