节点文献
通过正则化实现整周模糊度快速搜索
Quick Search of Ambiguity through Regularization
【摘要】 在讨论迭代双乔里斯基整数变换降相关性LAMBDA方法的基础上,发现解算历元数较少时方差-协方差阵轻微病态,提出了在分解前对方差-协方差阵正则化的改进法,实例证明改进后的方法需要历元数减少,搜索效率和稳定性高。
【Abstract】 This paper discusses the LAMBDA of iterative double-Cholesky integer transforming for decorrelation firstly,we found that the covariance matrix may be abnormity when a few epochs are used.So we introduce a modified method that the covariance matrix is regularized before decorrelating admissible integer transformation.The sample demonstrates that the modified method needs a smaller number of epochs and is more stable.
【关键词】 LAMBDA方法;
乔里斯基(Cholesky)分解;
Tikhonov正则化;
【Key words】 LAMBDA; Cholesky Decorrelation; Tikhonov Regularization;
【Key words】 LAMBDA; Cholesky Decorrelation; Tikhonov Regularization;
- 【文献出处】 全球定位系统 ,Gnss World of China , 编辑部邮箱 ,2009年03期
- 【分类号】TN967.1
- 【被引频次】9
- 【下载频次】185