Multiple GNSS Ambiguity Resolution And Validation In Short Baseline Positioning

The development of GNSS multi-frequency technology has greatly improved the performance of satellite navigation system,extending the scope of GNSS navigation and positioning services.The paper's main contents include multi-frequency combination observation theory,data quality control method,integer ambiguity resolution and ambiguity validity test based on multi-frequency GNSS data processing theory.The main work and research results are as follows:(1)Attempted to use the measured data to quality control for determine the weight.Firstly,the no-geometrical and no-ionospheric combination is used to eliminate the system error,and the noise residuals of the pseudorange and phase are separated.The noise errors are solved for each satellite by using multiple epochs.Finally,the method is verified by the Beidou triple-frequency measured data.(2)Focused on the ambiguity partial solution,and explored a new method of partial solution according to the difficulty of every ambiguity in the actual data.Compared and analyzed the advantages and disadvantages of the existing ambiguity resolution methods,included CIR,geometric CIR,LAMBDA,partial solution based on height angle and so on.Because of the problem that partial solution is not enough to locate,so the paper explores a more effective method of partial ambiguity solution.The method is used all satellites'information,and based on the difficulty of fixing the ambiguity in the actual data,and solves the problem that the partial ambiguity is too small to locate.Compared with all solution and the partial solution based on height angle,the results show that the fixed rate of the new method is 100%,and the numbers of fixed ambiguities are more than 15 in each epoch,in addition,the error rate of the new method is the lowest.(3)Proposed a new ambiguity validity test method called the conditional success rate.Researched the method of discriminant test,integer estimation method,integer aperture estimation method and the ratio test of controllable failure rate,then,the least squares success rate of the integer estimation method is discussed emphatically.Since the least square success rate cannot reflect the reliability of the actual ambiguity fixed,the paper proposed the conditional success rate method.This method takes into account the different position of the float point in the pull-in region,the closer to the center of the pull-in region,the higher the conditional success rate.Using the dual-frequency GPS data,it was shown that the conditional success rate method improved the reliability of ambiguity fixed,especially when the value of R-ratio was 1 nearly.