BeiDou Reliable Integer Ambiguity Resolution in the Presence of Time-Correlated Obversion Errors

摘要

Time to first fix of ambiguity is always finished by using multiple sequential epochs in real-time kinematic (RTK). Among different epochs, the residual observation errors exist the time correlation each other, which increases with the increase of distance of baseline. However, the characteristics of time correlation is always neglected by traditional RTK stochastic model to limit the performance of ambiguity resolution. In this paper, the characteristics of time correction among different epochs with BeiDou observations are analyzed. The magnitudes of the time correction from GEO satellites of BeiDou show larger than the MEO and IGSO satellites of BeiDou and all MEO satellites of GPS. Therefore, in this paper, the characteristics of time correction are modeled by using first-order Gaussian Markov processing to further adjust the RTK stochastic model. The proposed method is tested by a Monte-Carlo simulation and actual BeiDou observations. The simulated test shows that the proposed method can get more real probability of correct fix, compared with the model neglecting the characteristics of time correction. The actual test show that the proposed method deceases the probability of false alarm and missed detection by 50 and 15%, respectively, compared with the model neglecting the characteristics of time correction. As a result, the proposed method achieves the optimal ambiguity resolution performance taking the probability of success, false alarm and missed detection into account, simultaneously.