改进LAMBDA算法的双频BDS整周模糊度快速固定

Rapid resolution of dual frequency BDS integer ambiguity by improved LAMBDA algorithm

针对双频BDS短基线解算法方程病态严重,模糊度的浮点解精度低,导致LAMBDA算法搜索范围过大、搜索时间过长等问题,提出一种改进的LAMBDA算法:运用Tikhonov正则化理论构造一种具有明确物理意义的正则化矩阵R,以获得精确的整周模糊度浮点解;同时用均方误差矩阵RMSEM代替原算法的协方差阵确定模糊度的搜索范围,缩短初始化时间,提高搜索效率。实验结果表明:改进后的LAMBDA算法获得的模糊度浮点解精度显著提高,且无需初始化时间,能够正确固定模糊度的整数解,快速实现cm级定位。

Aiming at the problem that the solution algorithm of the BDS dual frequency short baseline has ill conditioned equation,and the floating point solution of ambiguity has low precision,leading to the large searching scope and long searching time in using LAMBDA algorithm,the paper proposed an improved LAMBDA algorithm：Tikhonov regularization theory was used to construct a regularized matrix R with a clear physical meaning for obtaining the accurate floating point solution of integer ambiguity;meanwhile a mean square error matrix RMSEM was used to replace the covariance matrix of the original algorithm in order to determine the searching range of ambiguity,thus the initialization time was reduced and the searching efficiency was improved.Experimental result showed that by the proposed method,the precision of the floating point solution of ambiguity could be increased significantly without initialization time,the integer ambiguity could be fixed correctly and the positioning with the centimeter level accuracy could be rapidly realized.