科研之友
微信
新浪微博
Facebook
分享链接
摘要
As the number of satellites increases, the risk of fixing wrong integer ambiguities may reduce the accuracy and efficiency of ambiguity resolution (AR); Thus, partial ambiguity resolution (PAR) is proposed for solving full AR (FAR) based on fixing the subset of ambiguities. Batch PAR with integer least-squares (ILS) is helpful, but selecting an optimal subset must compare all possible combinations, which is time-consuming. In this study, we introduced the classic optimal stopping theory (OST) to dynamically identify ambiguity subsets in batch PAR, which aims to maintain the accuracy and reduce the process time. The process of selecting the ambiguity subset was divided into two stages: observation and decision. All the options in the observation stage were discarded, and the choice was only made in the decision stage. The classic OST has demonstrated that there is a maximum 37% probability of selecting an optimal option when the range of observations is 37% based on six assumptions. To better explain the principle of OST in subset selection and provide a reference method in batch PAR, we chose the ambiguity dilution of precision (ADOP) PAR as the subject and put forward the OST-assisted ADOP PAR (OAPAR) to obtain the fixed solutions. The static relative positioning experiment began with reproducing a 37% observation range in OST (OST-37%) using the global positioning system (GPS) L1 static observation. The 37% observation range has a maximum probability of 41% in selecting the subset with the minimum ADOP and has less consumed time compared to other larger observation ranges; Then, FAR and minimum ADOP PAR (MAPAR) models are used to verify the accuracy and efficiency of OAPAR. The static results of different system combinations show that, compared to MAPAR, OAPAR could achieve the same accuracy while saving consumed time, especially when the number of ambiguities is higher. Finally, PAR based on success rate (SRPAR) is used to analyze the advantages and disadvantages of OAPAR. Although the probability of correct fixing of OAPAR is slightly smaller than SRPAR, the convergence time of OAPAR is better than SRPAR. Meanwhile, with the number of ambiguities increase, the RMSEs of OAPAR gradually perform smaller than SRPAR. Furthermore, the classic OST proposes a solution to the extreme value problem in global navigation satellite system (GNSS) positioning and navigation.
