Integer Ambiguity Resolution For GNSS-RTK Positioning In Complex Monitoring Environments

With the completion of the global network of China’s BD3 satellites,the Beidou navigation satellite system has officially provided global position services,and improved the high-precision,availability and reliability of GNSS navigation and positioning.Among them,the ground enhancement system with Real Time Kinematic(RTK)as the core has made some progress and development.However,satellite signals are easily obstructed and interfered,especially when the newly launched satellites signal is weak and rough values are prone to appear,and the ambiguity resolution between reference stations will be greatly affected.Moreover,with the integrative development of GNSS multi systems,multi frequencies and multi modes,conventional ambiguity algorithms are difficult to complete ambiguity search and resolution in high-dimensional scenarios.Study on the theory and method of GNSS-RTK positioning and ambiguity resolution is not only of great significance for high-precision positioning services under multiple GNSS conditions,but also conducive to the industrial promotion of BDS in the field of high-precision monitoring.This paper conducts in-depth research on the ambiguity resolution method for GNSS-RTK positioning in complex monitoring scenarios,focusing on creating the estimation model of noise reduction and robust ambiguity float solution for network RTK reference stations,as well as proposing the ambiguity resolution method for selecting subsets of medium and long baselines with condition constraints and even double-layer constraints,and the fast search method for high-dimensional integer ambiguities.The algorithms are verified and analyzed using measured data,and applied to a geological hazard monitoring system.The main work and innovation points are as follows:(1)Aiming at the problem of significant loss and low accuracy in the multipath error extraction model of network RTK reference stations,the estimation model of combined noise reduction and robust for ambiguity float solution between the reference stations is created.A multi-scale anti multipath joint denoising algorithm is designed,which extract the multipath error in the residual amount of carrier phase observation,and analyze the correlation of adjacent periods according to sidereal filtering(SF),and determine the time shift correction amount,and introduce robust model M-means estimation into the extended Kalman filtering algorithm,so as to improve the positioning accuracy of reference station baseline float solution.(2)Aiming at the problem of weak signal,long initialization time for ambiguity fixation,and slow convergence speed of newly launched satellites,the ambiguity resolution method for selecting subsets with condition constraints for medium and long baselines is proposed an improved.According to systematic deviations such as atmospheric delay corresponding to low elevation satellites,a method for selecting optimal subset of partial ambiguities with condition constraints is proposed to achieve fast baseline ambiguity resolution assisted by ionospheric weighted strategy.The verification shows that the algorithm has improved the ambiguity fixation rate by 16.6% and 6.4% respectively,compared to the FAR algorithm and the PAR one.(3)Aiming at the problem such as long initialization and slow convergence of high-dimensional ambiguity resolution,the method for selecting subsets with double-layer constraints in high-dimensional scenarios is proposed.An optimal subset selection strategy is proposed,which combines the optimal subsets at the satellite level and ambiguity level to achieve fast ambiguity resolution between network RTK reference stations.The verification shows that the algorithm has a 19.1% and 8.9% improvement in ambiguity fixation rate compared to FAR algorithm and PAR one respectively.(4)Aiming at the problem that the common ambiguity algorithm in the GNSS multi frequency,multi mode,and high-dimensional scenarios is difficult to meet ambiguity fixation,a fast search method for high-dimensional ambiguities based on lattice theory is proposed.This work investigates the effective equivalence problem between the least squares method based ambiguity search method and the lattice theory based shortest lattice point search method,and transforms the ambiguity search into the shortest distance lattice search.A Breadth-first traversal ambiguity search algorithm is proposed based on the lattice specification to maintain a high resolution fixation rate in high-dimensional ambiguity scenarios.The experimental results show that the Breadth-first search algorithm is more efficient in searching for integer solutions in high-dimensional ambiguities.