系统工程与电子技术 ›› 2023, Vol. 45 ›› Issue (9): 2860-2865.doi: 10.12305/j.issn.1001-506X.2023.09.25

• 制导、导航与控制 • 上一篇    下一篇

MPOS联邦实时组合算法

李建利, 魏梦笛, 王其朋, 张武   

  1. 北京航空航天大学仪器科学与光电工程学院, 北京 100191
  • 收稿日期:2022-05-28 出版日期:2023-08-30 发布日期:2023-09-05
  • 通讯作者: 魏梦笛
  • 作者简介:李建利(1979—), 男, 教授, 博士, 主要研究方向为微小型惯性导航、高精度惯性导航、惯性导航系统/全球导航卫星系统组合导航系统技术
    魏梦笛(1997—), 女, 硕士研究生, 主要研究方向为微小型惯性导航
    王其朋(1995—), 男, 博士研究生, 主要研究方向为微小型惯性导航
    张武(1997—), 男, 硕士研究生, 主要研究方向为微小型惯性导航

MPOS federated real-time integration method

Jianli LI, Mengdi WEI, Qipeng WANG, Wu ZHANG   

  1. School of Instrumentaton and Optoelectronic Engineering, Beihang University, Beijing 100191, China
  • Received:2022-05-28 Online:2023-08-30 Published:2023-09-05
  • Contact: Mengdi WEI

摘要:

微小型位置姿态测量系统(micro position and orientation system, MPOS)是为成像载荷提供实时高精度运动补偿信息的关键装置, 其测量精度严重制约成像精度的提升。针对MPOS集中滤波实时性差的问题, 在基于双ARM (advanced reduced instruction set computing machines)硬件平台的基础上, 采用联邦滤波实时组合算法对多组传感器数据进行最优信息融合。以微惯性测量单元为公共参考系统, 双天线全球导航卫星系统(global navigation satellite system, GNSS)和磁强计作为子系统提供量测信息, 将各个子系统得到的局部误差协方差及状态估计值在主滤波器中进行信息融合得到最优估计值。通过动态实验验证了基于联邦滤波实时组合方法的MPOS, 其位置、速度、航向角及水平姿态角精度可达0.6 m, 0.03 m/s, 0.15°, 0.04°。

关键词: 微小型位置姿态测量系统, 联邦滤波, 实时组合, 全球导航卫星系统, 磁强计

Abstract:

Micro position and orientation system (MPOS) is a key device to provide real-time and high-precision motion compensation information for imaging loads, and its measurement accuracy severely restricts the improvement of imaging accuracy. Aiming at MPOS real-time integrated filtering algorithm, based on dual-ARM (advanced RISC machines) hardware platform, this paper adopts federated filtering real-time integrated method to achieve optimal information fusion of multiple groups of sensor data. Micro inertial measurement unit is the common reference system, double antenna global navigation satellite system(GNSS) and magnetometer are the measuring information of subsystem. The local error covariance and state estimation obtained by each subsystem are fused in the main filter to obtain the optimal estimation value. The MPOS system based on real-time federated filtering integration method is verified by dynamic experiments. The accuracy of position, speed, heading and horizontal attitude can reach 0.6 m, 0.03 m/s, 0.15° and 0.04°.

Key words: micro position and orientation system (MPOS), federated filtering, real-time integration, global navigation satellite system (GNSS), magnetometer sensor

中图分类号: