航空学报 > 2014, Vol. 35 Issue (6): 1665-1672   doi: 10.7527/S1000-6893.2013.0544

无需中间变量的多运动站时差定位新算法

徐征, 曲长文, 骆卉子   

  1. 海军航空工程学院 电子信息工程系, 山东 烟台 264001
  • 收稿日期:2013-08-21 修回日期:2014-02-18 出版日期:2014-06-25 发布日期:2014-02-28
  • 通讯作者: 曲长文,Tel.:0535-6635080 E-mail:qcwwby@sohu.com E-mail:qcwwby@sohu.com
  • 作者简介:徐征男,博士研究生。主要研究方向:无源定位跟踪技术。Tel:0535-6635695 E-mail:xuzheng85@126.com;曲长文男,博士,教授,博士生导师。主要研究方向:无源定位跟踪技术、SAR成像技术、SAR目标检测与识别技术。Tel:0535-6635080 E-mail:qcwwby@sohu.com;骆卉子女,博士研究生。主要研究方向:无源定位跟踪技术。Tel:0535-6635695 E-mail:jessica_lhz@163.com
  • 基金资助:

    航空科学基金(20105584004)

Novel Multiple Moving Observers TDOA Localization Algorithm Without Introducing Intermediate Variable

XU Zheng, QU Changwen, LUO Huizi   

  1. Department of Electronic and Information Engineering, Naval Aeronautical and Astronautical University, Yantai 264001, China
  • Received:2013-08-21 Revised:2014-02-18 Online:2014-06-25 Published:2014-02-28
  • Supported by:

    Aeronautical Science Foundation of China(20105584004)

摘要:

传统时差 (TDOA)定位模型通过引入中间变量来得到线性方程,需要两步求解过程且该模型不适合多运动站连续定位。为此,引入无需中间变量的时差定位模型,并在此基础上提出了一种约束加权最小二乘定位算法。首先将基于该模型的时差定位问题转换为加权最小二乘问题,然后推导代入时差测量值后观测矩阵和观测向量的误差项,将其每一列表示为确定矩阵与随机时差测量噪声向量乘积的形式,并基于此推导了关于目标状态的二次约束方程,最终只需通过广义特征值分解来得到目标状态估计,并推导了该估计的解析表达式。仿真结果表明所提算法的连续定位性能逼近克拉美罗-限且所得定位解渐近无偏。

关键词: 定位, 时差, 约束加权最小二乘, 广义特征值分解, 克拉美罗-限

Abstract:

The traditional time difference of arrival (TDOA) localization model needs an intermediate variable to obtain the linear equation, which requires a two-step solution procedure and is not suitable for multiple moving observers continuous localization. Therefore, a TDOA localization model without the intermediate variable is introduced and a constrained weighted least squares algorithm is proposed based on it. First, a weighted least squares problem is formed with respect to the model. Then, error items are deduced after substituting TDOAs in the observation matrix and the observation vector for the measured ones. Each column of error items is expressed as the product of a deterministic matrix and a random vector composed of TDOA measurement errors, based on which the quadratic constraint on the target state is formed. Finally, the estimated target state is obtained through generalized eigendecomposition and its analytic form is derived. Simulation results indicate that the proposed algorithm achieves the Cramer-Rao lower bound and has an asymptotically unbiased solution.

Key words: localization, time difference of arrival, constrained weighted least squares, generalized eigendecomposition, Cramer-Rao lower bound

中图分类号: