系统工程与电子技术 ›› 2019, Vol. 41 ›› Issue (4): 730-736.doi: 10.3969/j.issn.1001-506X.2019.04.05

• 电子技术 • 上一篇    下一篇

基于机动判别的变结构交互多模型跟踪算法

潘媚媚, 曹运合, 王宇, 吴文华   

  1. 西安电子科技大学雷达信号处理国家重点实验室, 陕西 西安 710071
  • 出版日期:2019-03-20 发布日期:2019-03-20

Variable structure interactive multi-model tracking algorithm based on maneuvering discriminant

PAN Meimei, CAO Yunhe, WANG Yu, WU Wenhua   

  1. National Lab of Radar Signal Processing, Xidian University, Xi’an 710071, China
  • Online:2019-03-20 Published:2019-03-20

摘要:

临近空间高超声速机动目标具有高速、高机动的运动特性,协同转弯模型是跟踪临近空间目标的常用模型之一。基于协同转弯模型的自适应网格交互多模型(adaptive grid interaction multiple model, AGIMM)算法能够较好地适应临近空间高超声速目标运动特性,但AGIMM算法存在着依赖中心网格模型,非机动时刻模型集收敛缓慢的问题,基于此提出了一种基于机动判别的变结构交互多模型算法。所提算法根据目标机动特性调整跟踪模型集的结构及模型概率转移矩阵,加快了非机动时刻模型集的收敛速度,克服了AGIMM算法存在的问题。通过蒙特卡罗仿真验证,改进的算法相对于AGIMM算法提高了对临近空间高超声速目标的跟踪性能。

Abstract:

The near space hypersonic maneuvering targets have the motion characteristics of highspeed and highmobility.The collaborative turning model is one of the common models for tracking near space targets.The adaptive grid interaction multiple model (AGIMM) algorithm based on collaborative turning model can better adapt to the motion characteristics of the near space hypersonic targets.Aiming at the problem that AGIMM algorithm is dependent on the central grid model and the model set converges slowly at nonmaneuvering time,a variable structure interactive multi model algorithm based on maneuvering discriminant is proposed.The proposed algorithm adjusts the structure of the tracking model set and the model probability transfer matrix according to the maneuvering characteristics of the targets,which accelerates the convergence speed of the model set at nonmaneuvering time and overcomes the problems of the AGIMM algorithm.Monte Carlo simulations verify that the improved algorithm improves the tracking performance of hypersonic targets in the near space compared with the AGIMM algorithm.