欢迎访问《兵工学报》官方网站,今天是 分享到:

兵工学报 ›› 2015, Vol. 36 ›› Issue (3): 427-432.doi: 10.3969/j.issn.1000-1093.2015.03.007

• 论文 • 上一篇    下一篇

基于截断总体最小二乘算法的车载三轴磁力仪标定

张滢1, 杨任农1, 李明阳1, 左家亮1, 陈星伊2   

  1. (1.空军工程大学 航空航天工程学院陕西 西安 710051; 2.西安卫星测控中心, 陕西 西安 710043)
  • 收稿日期:2014-04-17 修回日期:2014-04-17 上线日期:2015-05-01
  • 作者简介:张滢(1988—)女博士研究生

Calibration of Vehicular Three-axis Magnetometer via Truncated Total Least Squares Algorithm

ZHANG Ying1, YANG Ren-nong1, LI Ming-yang1, ZUO Jia-liang1, CHEN Xing-yi2   

  1. (1.Aeronautics and Astronautic Engineering College, Air Force Engineering University, Xi’an 710051, Shaanxi, China;2.Xi’an Satellite Control Center, Xi’an 710043, Shaanxi, China)
  • Received:2014-04-17 Revised:2014-04-17 Online:2015-05-01

摘要: 以陆用车载磁导航为应用背景,提出一种正则化三轴磁力仪标定算法。由于车辆的机动受到限制,用于标定三轴磁力仪的观测方程是严重病态的。传统的三轴磁力仪标定方法缺乏对病态问题的处理,难以获得准确的标定参数。文中采用截断总体最小二乘(TTLS)技术对病态的磁力仪标定问题进行求解。该方法能够有效地抑制观测方程两端的误差,并且适合处理强病态问题。通过车载实验对该方法的有效性进行研究,结果表明文中提出的方法能够有效地减缓磁力仪标定中的病态影响,获得更为稳定的数值解,补偿后的平均航向角估计精度为1.5°,满足车载磁导航的精度要求。

关键词: 控制科学与技术, 截断总体最小二乘, 标定, 车辆导航

Abstract: A regularization three-axis magnetometer calibration algorithm is proposed for vehicular navigation. Due to the constrained manoeuvre of the vehicle, a linear observational equation system for calibration of three-axis magnetometer is severely ill-posed. The traditional calibration methods may result in unreliable solutions in the lack of dealing with ill-posed problem. The ill-posed problem is solved by using the truncated total least squares (TTLS) technique. This method can effectively suppress the errors on both sides of the observation equation. Furthermore, the TTLS method is used to solve the ill-posed problems. A vehicular navigation experiment is conducted to validate the effectiveness of the method.The results show that the proposed algorithm can effectively mitigate the ill-posed problem and obtain more stable numerical solutions. The mean value of the heading angles from the calibrated measurements is equal to 1.5°, which meets the vehicular navigation application.

Key words: control science and technology, truncated total least squares, calibration, vehicle navigation