系统仿真学报 ›› 2021, Vol. 33 ›› Issue (2): 315-323.doi: 10.16182/j.issn1004731x.joss.19-0502

• 仿真建模理论与方法 • 上一篇    下一篇

基于模糊自适应滤波的单轴旋转惯导精对准方法

胡杰1,2, 石潇竹1,2   

  1. 1.中国电子科技集团公司第二十八研究所,江苏 南京 210007;
    2.空中交通管理系统与技术国家重点实验室,江苏 南京 210007
  • 收稿日期:2019-09-06 修回日期:2019-11-09 出版日期:2021-02-18 发布日期:2021-02-20
  • 作者简介:胡杰(1987-),男,博士,工程师,研究方向为惯性导航及组合导航算法。E-mail:hj_student@163.com
  • 基金资助:
    国家重点研发计划(2017YFB0503401),江苏省自然科学基金青年基金(BK20170157)

Refined Alignment Method for Single-axis Rotary Inertial Navigation Based on Fuzzy Adaptive Filtering

Hu Jie1,2, Shi Xiaozhu1,2   

  1. 1. The 28th Research Institute of China Electronics Technology Group Corporation, Nanjing 210007, China;
    2. State Key Laboratory of Air Traffic Management System and Technology, Nanjing 210007, China
  • Received:2019-09-06 Revised:2019-11-09 Online:2021-02-18 Published:2021-02-20

摘要: 为了抑制单轴旋转惯导航系统中惯性测量单元旋转产生的“锯齿形”速度误差对导航系统初始对准精度影响,提出了一种基于模糊自适应Kalman滤波的精对准方法。该方法计算新息序列实际协方差与理论协方差比值后利用模糊推理系统自适应调节量测噪声协方差矩阵,使其能够适应惯性测量单元旋转而引起的量测噪声变化。进行了摇摆环境下验证实验,结果表明,自适应滤波算法能够有效抑制惯性测量单元旋转引起的误差,航向角误差由2.31°减小至0.2°,俯仰角误差由0.11°减小为0.02°,横摇角误差由0.09°减小为-0.03°,同时,转台不同方位下的初始对准精度满足高精度捷联惯导系统导航需求。

关键词: 捷联惯性导航系统, 单轴旋转, 自适应Kalman滤波, 模糊推理系统

Abstract: In order to restrain the influence of saw tooth velocity error caused by the inertial measurement unit (IMU) rotation on the initial alignment accuracy of single-axis rotary strapdown inertial navigation system (SINS), a refined alignment method based on fuzzy adaptive Kalman filtering is proposed. After calculating the ratio of actual covariance to theoretical covariance of the innovation sequence, the method adaptively adjusts the measurement noise covariance matrix by using the fuzzy inference system (FIS), so that it can adapt to the changes of measurement noise caused by IMU rotation. The initial alignment validation experiments under the swing environment have been carried out and experiment results show that the adaptive filtering algorithm can effectively restrain the filtering output error caused by the rotation of the IMU, and the heading angle error is reduced from 2.31° to 0.2°, the pitch angle error is reduced from 0.11° to 0.02°, and the roll angle error is reduced from 0.99° to -0.03°. Meanwhile, the initial alignment accuracy of the turntable in different azimuths meets the navigation requirements of high precision SINS.

Key words: strapdown inertial navigation system, single-axis rotary, adaptive Kalman filtering, fuzzy inference system

中图分类号: