一种改进两步加权最小二乘的TDOA定位算法  

An Improved Two-step Weighted Least Squares TDOA Location Algorithm

作　　者：孟天次 张贞凯[1] 林云航 MENG Tianci;ZHANG Zhenkai;LIN Yunhang(Ocean College,Jiangsu University of Science and Technology,Zhenjiang 212028,China)

机构地区：[1]江苏科技大学海洋学院,江苏镇江212028

出　　处：《电讯技术》2022年第6期782-787,共6页Telecommunication Engineering

基　　金：国家自然科学基金资助项目(61871203);江苏省研究生科研与实践创新计划(SJCX22_1886)。

摘　　要：现有的两步加权最小二乘水下声源定位算法中,两步法处理均存在平方以及开方等非线性运算忽略了二阶误差项,导致定位精度下降。提出了一种改进的两步加权最小二乘算法,首先引入时差方程,构建新的定位方程,初步估计出声源位置;接着引入中间变量并进行泰勒展开,利用中间变量与声源位置误差之间的非线性关系构建定位方程,对声源位置初步估计值进行校正,有效避免了二阶误差项的丢失,得到了更加精确的估计结果。通过与克拉美罗下界进行比较,验证了算法性能的同时进一步提升了算法定位精度。In the existing two-step weighted least squares underwater sound source localization algorithm,there are non-linear operations such as squaring and square root in the two-step processing,which ignore the second-order error term,resulting in a decrease in positioning accuracy.An improved two-step weighted least squares algorithm is proposed.First,the time difference equation is introduced,a new positioning equation is constructed,and the sound source position is preliminarily estimated.Then,the intermediate variable is introduced and Taylor expansion is performed,the non-linear relationship between the intermediate variable and the position error of the sound source is used to construct a positioning equation and correct the preliminary estimation value of the sound source position,effectively avoiding the loss of the second-order error term and obtaining a more accurate estimation result.Through comparion with the Cramer-Row lower bound,the performance of the algorithm is verified,and the positioning accuracy of the algorithm is further improved.

关 键 词：水下声源定位 到达时间差(TDOA) 改进最小二乘 二阶误差项 

分 类 号：TN971[电子电信—信号与信息处理]