航空学报 > 1999, Vol. 20 Issue (3): 234-238

联合交互式多模型概率数据关联算法

潘泉, 刘刚, 戴冠中, 张洪才   

  1. 西北工业大学自动控制系, 陕西西安 710072
  • 收稿日期:1998-05-11 修回日期:1998-07-15 出版日期:1999-06-25 发布日期:1999-06-25

COMBINED INTERACTING MULTIPLE MODELS PROBABILISTIC DATA ASSOCIATION ALGORITHM

PAN Quan, LIU Gang, DAI Guan-zhong, ZHANG Hong-cai   

  1. Department of Automatic Control, Northwestern Polytechnical University, Xi'an 710072, China
  • Received:1998-05-11 Revised:1998-07-15 Online:1999-06-25 Published:1999-06-25

摘要:

通过对交互式多模型(IMM)算法和概率数据关联(PDA)算法的研究,指出Bar-Shalom和Blom等提出的IMMPDA算法结构和理论上存在的问题。根据全局最优的思想,设计了一个最优波门,从而得到全局最优的量测集合,保证了系统的理论完整性和结构合理性。推导完成了新的联合IMMPDA算法——C-IMMP-DA算法。大量仿真计算验证了C-IMMPDA算法在减少计算量的同时,总体性能上较之原IMMPDA算法有大幅度提高。

关键词: 交互式多模型, 概率数据关联, 多目标跟踪

Abstract:

Considering the four defects of construction and theory of IMMPDA algorithm which is proposed by Y. Bar Shalom and H. A. P. Blom, the authors work out a new Combined IMMPDA algorithm C IMMPDA algorithm. In C IMMPDA, a global optimal validating gate for every sub filter is constructed instead of the different gates for every sub filter in IMMPDA algorithm. That means it has only one PDAF in the filter system. All the formulas are deduced and the diagram of the new algorithm is drawn. The work makes the IMMPDA more satisfactory in construction and theory. Simulation results show that the new algorithm will decrease the computational burden and be of much better performance than IMMPDA.

Key words: interact ing mult ip le models algo rithm, p robabilist ic data associat ion algo rithm, mult ip le targetst rack ing

中图分类号: