| Recently,with the development of GPS,many researchers pay attention to the aircraft's precise entrance and landing through GPS carrier phase measurement. The key technology of aircraft's precise entrance and landing is GPS carrier phase integer ambiguity resolution on the fly. So,in this paper,the problems in this field are studied and the improving algorithm is proposed and simulated. And then,the statistics of integer ambiguity estimation are analyzed;the method of improving the success probability of integer ambiguity estimation is presented. The research and consequence are as follows.1) Based on the theory and application of GPS,the construction and application of the RTK (Real Time Kinematic)-GPS carrier phase measurement (so called RTK) are studied. The differences between messages type 18/19 and type 20/21 are compared by analysing RTCM SC-104 messages,and the advantages and disadvantages of using those types are pointed out.2) The various factors,which influence positioning accuracy,are analyzed in this paper. The effect of multipath,which can not be removed through differential GPS positioning,is studied in theory exceptionally. At the same time,the DGPS and RTK are constructed,and a lot of DGPS and RTK positioning data are collected. Analysis in theory and Data processing show that the DGPS positioning error comes from the effect of multipath mostly,and RTK positioning error comes from the effect of multipath and the receiver noise.3) The GPS integer ambiguity estimates,the navigation parameter estimates and their variance-covariance matrix are derived by using Markovian estimation. The GPS pseudorang observation equation and carrier phase observation equation are constructed and linearized. The LAMBDA is regarded as a more effective algorithm for realizing AROF (Ambiguity Resolution On the Fly) after analyzing and comparing a lot of GPS integer ambiguity resolution algorithms. The integer least squares estimation of the LAMBDA is analyzed,and the result shows that the key process of integer least squares estimation is GPS ambiguity decorrelation.4) Based on the matrix theory and the constraint condition of the Z transformation,a rapid and effective algorithm of ambiguity decorrelation is provided. This algorithm is called OSD (ordering screen decorrelation). The ordering process,the integer LDLT factorization and the ambiguity screen are three parts of the OSD. The OSD is based on diminishing the condition number of covariance matrix. Compared with the current whitening transformation,it can get the same result,but the OSD operational time isabout half of the whitening transformation's when it processes the same ambiguity covariance matrix. And the ambiguity screen can improve the reliability of the ambiguity decorrelation.5) By analyzing and demonstrating the random characteristics of GPS integer ambiguity estimation,we show that GPS integer ambiguity estimation is depended on mode of estimation. Then,we analyze the distributed characteristics and estimation performance of three typical estimation algorithms,and point out that integer least squares algorithm has the biggest success probability. The approximate computing method of the integer least squares success probability is presented. By investigating the three modes of navigation and positioning parameter estimation,we find and prove that they have different mean and variance. At last,we study the fixed estimation statistics of navigation and positioning parameter,compare positioning accuracy of the condition estimation with that of the fixed estimation,and point out interrelation of them.6) After investigating the two methods of increasing the success probability,a new method of increasing success probability is presented. Based on the quality of ambiguity estimators,the new method carries out the ambiguity ordering process,chooses ambiguities by estimators' accuracy,and goes on ambiguity decorrelation process using those chosen ambiguities with high accuracy. By mathematical analysis and simulation,it can be proven t...
|
PDF Full Text Request