Scholarlink

The problem of state estimation accuracy decline caused by model uncertainty in the field of target tracking, this paper first presents an IMM nonlinear filter based on the Gaussian framework, uses IMM to initialize the effective model, uses the high-order UT method to construct Sigma points and weights to estimate the state random variables, and proposes an IMM nonlinear filter based on the high-order UT framework. Compare and analyze performance indicators under different filters using a maneuvering target tracking model. The simulation results show that the position error of IMMHUKF is 95% and 48% higher than that of traditional EKF algorithm and UKF algorithm, and 97% and 53% higher than that of EKFIMM and UKFIMM algorithm; velocity error: Compared with the traditional EKF algorithm and UKF algorithm, IMMHUKF has 75% and 63% higher precision, and 81% and 15% higher precision than EKF-IMM and UKF-IMM algorithms. Thus, IMMHUKF can obtain higher precision of state estimation.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A New Interactive Multi Model High Order Unscented Kalman Filter for Improving Target Tracking Accuracy and Robustness

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Jiaxin Hou,
  • Houpu Li,
  • Shuguang Wu,
  • Yanting Yu,
  • Juntong Chen,
  • Meng Li,
  • Shaofeng Bian

摘要

The problem of state estimation accuracy decline caused by model uncertainty in the field of target tracking, this paper first presents an IMM nonlinear filter based on the Gaussian framework, uses IMM to initialize the effective model, uses the high-order UT method to construct Sigma points and weights to estimate the state random variables, and proposes an IMM nonlinear filter based on the high-order UT framework. Compare and analyze performance indicators under different filters using a maneuvering target tracking model. The simulation results show that the position error of IMMHUKF is 95% and 48% higher than that of traditional EKF algorithm and UKF algorithm, and 97% and 53% higher than that of EKFIMM and UKFIMM algorithm; velocity error: Compared with the traditional EKF algorithm and UKF algorithm, IMMHUKF has 75% and 63% higher precision, and 81% and 15% higher precision than EKF-IMM and UKF-IMM algorithms. Thus, IMMHUKF can obtain higher precision of state estimation.