改进的整周模糊度搜索算法  

An improved method for searching integer ambiguity

作　　者：宋福成[1] 杨汀[1] 陈宜金[1] 时爽爽 

机构地区：[1]中国矿业大学(北京),北京100083 [2]武汉大学,武汉430079

出　　处：《测绘科学》2015年第10期16-20,共5页Science of Surveying and Mapping

基　　金：中央高校基本科研业务费项目(2010YD06)

摘　　要：针对现有模糊度搜索方法仍不能很好地满足快速定位需求的问题,该文在简要介绍解决最近向量问题的搜索算法基础上,将M-VB搜索算法引入到模糊度的解算中,并对其作了两方面改进。一是优化了该算法执行过程中更新上界的问题,二是提出借助序贯最小二乘平差(Bootstrapped)估计值来确定其搜索空间半径的方法。基于仿真数据和实测GPS数据,分别在降相关和不降相关条件下,将上述改进方法与最小二乘降相关平差(LAMBDA)方法和其修正方法(MLAMBDA)作了对比分析。结果表明,改进的M-VB算法比其他2种方法能更快地固定整数向量,有效地提高了模糊度搜索效率。According to the fact that the current methods for ambiguity estimation do not meet with the demand of fast positioning, M-VB method was introduced for ambiguity estimation after presentation of searching methods for closest vector problem（CVP）. Then, M-VB method was improved from two points. One was optimizing the method with updating the upper bounds, and the other was determining the radius with bootstrapped estimator. The improved M-VB method was compared to LAMBDA and MLAMBDA methods based on simulated and real GPS data on the condition of with and without decorrela- tion. The results indicated that the M-VB method could fix integers faster than LAMBDA and MLAMBDA methods, and the searching efficiency was improved.

关 键 词：全球导航卫星系统 模糊度估计 最近向量问题 LAMBDA MLAMBDA M-VB 

分 类 号：P228.4[天文地球—大地测量学与测量工程]