| 引用本文: | 杨金一,郭妍,江鹏飞.稀疏基站条件下基于迭代卡尔曼滤波的协同定位方法[J].控制理论与应用,2023,40(12):2209~2216.[点击复制] |
| YANG Jin-yi,GUO Yan,JIANG Peng-fei.Cooperative positioning method based on iterated Kalman filter in sparse-beacon environments[J].Control Theory and Technology,2023,40(12):2209~2216.[点击复制] |
|
| 稀疏基站条件下基于迭代卡尔曼滤波的协同定位方法 |
| Cooperative positioning method based on iterated Kalman filter in sparse-beacon environments |
| 摘要点击 757 全文点击 256 投稿时间：2023-04-25 修订日期：2023-12-19 |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| DOI编号 10.7641/CTA.2023.30272 |
| 2023,40(12):2209-2216 |
| 中文关键词 协同定位 惯性导航系统 距离测量 量测迭代 |
| 英文关键词 cooperative positioning inertial navigation system range measurement measurement iteration |
| 基金项目 国家自然科学基金项目(62103424)资助. |
|
| 中文摘要 |
| 近年来, 多无人平台协同定位应用广泛, 各平台利用距离测量等相互观测可以进一步提高自身的定位精度.
在惯性/测距组合的协同定位系统中, 一般需要环境中存在4个以上位置已知的基站提供稳定测距, 通过各平台的相
互测距来修正惯导解算积累的定位误差. 但是, 在一些通信距离过大或者有遮挡的环境中, 平台不能准确接收来自
足够多基站的测距, 常用的基于误差状态扩展卡尔曼滤波的协同定位方法精度较差. 本文提出一种基于迭代卡尔
曼滤波的惯性/测距协同定位方法, 在误差状态扩展卡尔曼滤波的基础上, 以测距精度为阈值进行滤波更新过程中
误差状态的迭代计算, 减小扩展卡尔曼滤波省略高阶泰勒展开项所引起的非线性误差, 进而提高参与协同的各节点
定位精度. 稀疏基站场景下的仿真验证了本文所述方法的有效性. |
| 英文摘要 |
| In recent years, the cooperative positioning of multiple unmanned platforms has been widely applied, and each
platform can further improve its positioning accuracy by using range measurements and other mutual observations. In the
inertial/ranging combined cooperative positioning system, it is generally necessary to have four or more known base stations
in the environment to provide stable ranging, and to correct the accumulated positioning error of inertial navigation solution
through mutual ranging of each platform. However, in some environments with large communication distances or occlusion,
the range measurements received by the platforms are sparse, and the commonly used cooperative positioning method based
on the error state extended Kalman filter has poor accuracy. This paper proposes an inertial/ranging cooperative positioning
method based on the iterated Kalman filter. Based on the extended Kalman filtering for error states, the iterative calculation
of error states during the filtering update process is carried out with ranging accuracy as the threshold, reducing the nonlinear
error caused by the omission of high-order Taylor expansion terms by the extended Kalman filtering, and thereby improving
the positioning accuracy of each node participating in the cooperation. Simulation and physical experiments have verified
the effectiveness of the proposed method in different sparse-ranging environments. |
|
|
|
|
|