系统工程与电子技术 ›› 2021, Vol. 43 ›› Issue (7): 1824-1830.doi: 10.12305/j.issn.1001-506X.2021.07.13

• 电子技术 • 上一篇    下一篇

基于修正的自适应平方根容积卡尔曼滤波算法

李春辉1, 马健1, 杨永建1,2,*, 肖冰松1, 邓有为1, 盛涛1   

  1. 1. 空军工程大学航空工程学院, 陕西 西安 710038
    2. 西北工业大学电子信息学院, 陕西 西安 710072
  • 收稿日期:2020-09-01 出版日期:2021-06-30 发布日期:2021-07-08
  • 通讯作者: 杨永建
  • 作者简介:李春辉 (1997—), 男, 硕士研究生, 主要研究方向为雷达信号与信息处理|马健 (1972—), 男, 副教授, 硕士研究生导师, 博士, 主要研究方向为雷达信号与信息处理|杨永建 (1988—), 男, 讲师, 博士, 主要研究方向为雷达信号与信息处理|肖冰松 (1982—), 男, 副教授, 博士, 主要研究方向为航空武器系统总体设计|邓有为 (1981—), 男, 讲师, 硕士, 主要研究方向为目标跟踪算法|盛涛 (1995—), 男, 硕士研究生, 主要研究方向为雷达信号处理
  • 基金资助:
    空军工程大学校长基金(XZJ2020039)

Adaptive square-root cubature Kalman filter algorithm based on amending

Chunhui LI1, Jian MA1, Yongjian YANG1,2,*, Bingsong XIAO1, Youwei DENG1, Tao SHENG1   

  1. 1. Aeronautics Engineering College, Air Force Engineering University, Xi'an 710038, China
    2. School of Electronics and Information, Northwestern Polytechnical University, Xi'an 710072, China
  • Received:2020-09-01 Online:2021-06-30 Published:2021-07-08
  • Contact: Yongjian YANG

摘要:

目标建模不确定性会造成滤波算法性能下降, 通过构建强跟踪滤波器(strong tracking filter, STF)可以提升滤波算法的自适应性, 但是构建STF时存在理论推导复杂、求解计算量大等局限和不足, 针对上述问题, 在平方根容积卡尔曼滤波(square-root cubature Kalman filter, SRCKF)的基础上, 提出一种基于修正的自适应SRCKF算法。该算法通过设置判定门限和修正准则, 直接对状态预测值或滤波增益进行修正以平衡先验的预测值和后验反馈的量测值在滤波中所占的比重, 进而减小状态估计误差。仿真结果表明, 所提算法具有在目标状态突变和量测非线性时的良好滤波性能和数值稳定性, 同时相比较需要计算渐消因子的STF算法, 该算法在计算量和收敛速度上具有优势。

关键词: 目标建模, 平方根容积卡尔曼滤波, 修正算法, 自适应滤波

Abstract:

The uncertainty of target modeling will lead to the performance degradation of the filter algorithm, and the self-adaptability of the filter algorithm can be improved by constructing strong tracking filter (STF). However, there are limitations and deficiencies in the construction of STF, such as complex theoretical derivation and large amount of calculation. To solve the above problems, an adaptive square-root cubature Kalman filter (SRCKF) algorithm based on amending is proposed which is based on SRCKF. By setting judgment threshold and amending rules, the proposed algorithm directly amends the predicted state value or filter gain to balance the proportion of the predicted prior value and the measured posterior feedback value in the filtering, which can reduce the state estimation error. Simulation results show that the algorithm has good filtering performance and numerical stability when the target state is suddenly changed and the measurement is nonlinear. Meanwhile, compared with the STF algorithm which needs to calculate the fading factor, the proposed algorithm has advantages in calculation amount and convergence speed.

Key words: target model, square-root cubature Kalman filter (SRCKF), amending algorithm, adaptive filtering

中图分类号: