中文摘要：

利用北斗卫星导航系统进行基线解算时，正确解算模糊度是关键问题．利用北斗卫星导航系统双频载波相位观测值组成宽巷双差观测值，运用目前被认为效果最好的模糊度解算方法——LAMBDA算法解算宽巷双差整周模糊度，但是，该算法存在模糊度浮点解偏差大，搜索范围大的缺点．因此，该文提出利用Tikhonov正则化改进LAMBDA算法，对宽巷双差观测方程中未知参数的系数矩阵进行奇异值分解，用分解后的协方差矩阵替换LAMBDA算法中的协方差矩阵进行宽巷双差整周模糊度的搜索，提高了宽巷双差整周模糊度浮点解的精度和模糊度固定正确率．改进后的LAMBDA算法模糊度浮点解偏差平均值从1．0829周减少到0．0059周，方差值从4．0534周减少到0．0479周，模糊度固定正确率从93．52％提高到100％．

英文摘要：

Using the BeiDou Navigation Satellite System （BDS） for baseline settlement, the key problem is the correct solution of ambiguity. Using carrier phase observations of BDS to make up wide lane double difference observations, the best effect algorithm of LAMBDA is used to calculate wide lane double difference integer ambiguity. However, the algorithm has the disadvantage of large deviation of ambiguity float solution and large search range. So, the paper put forward to use Tikhonov to improve LAMBDA algorithm, singular value decomposition of the coefficient matrix of the unknown parameters in the wide lane double difference observation equation, replacing the covariance matrix of the LAMBDA algorithm with the decomposition of the covariance matrix and searching for the ambiguity of wide lane double difference, the precision of the float solution of the wide lane double difference integer ambiguity and the accuracy of ambiguity are improved. Deviation average values of ambiguity float solution of improved LAMBDA algorithm were reduced from 1.0829 cycles to 0.0059 cycles, variance values were reduced from 4.0534 cycles to 0.0479 cycles, the accuracy of the degree is raised from 93.52% to 100%.