电子学报

• 科研通信 • 上一篇    下一篇

交互式多区域模型

曲彦文1,2, 张二华2, 杨静宇2   

  1. 1. 江西师范大学计算机信息工程学院,江西南昌 330000;
    2. 南京理工大学计算机科学与技术学院,江苏南京 210094
  • 收稿日期:2011-02-18 修回日期:2011-09-10 出版日期:2012-06-25 发布日期:2012-06-25
  • 作者简介:曲彦文 男,1983年4月出生于江西省南昌市.现为江西师范大学讲师,从事信号处理、数据融合方面的研究. E-mail:earverse@hotmail.com
    张二华 男,1967年3月出生于湖北省武孝感市.现为南京理工大学副教授,从事科学可视化、信号处理方面的研究. E-mail:zherhua@163.com
  • 基金资助:

    国家自然科学基金(No.90820306)

Interacting Multiple Region Model

QU Yan-wen1,2, ZHANG Er-hua2, YANG Jing-yu2   

  1. 1. School of Computer Information and Engineer,Jiangxi Normal University,Nanchang,Jiangxi 330000,China;
    2. School of Computer Science and Technology,Nanjing University of Science and Technology,Nanjing,Jiangsu 210094,China
  • Received:2011-02-18 Revised:2011-09-10 Online:2012-06-25 Published:2012-06-25

摘要: 一种被称为交互式多区域模型(IMRM)的非线性滤波算法被提出,用于对状态和连续系统参数进行联合估计.IMRM将连续的系统参数空间视为由若干子区域所构成的集合,并将每个子区域分别分配给一个子模型.IMRM使用一组子滤波器并行滤波.在每一时刻,IMRM利用交互操作计算各子模型的混合初始化环境,之后各子滤波器在假设系统参数跳变到特定子区域的前提下,对状态和系统参数进行估计.为了有效地应用IMRM,提出了一种基于无迹变换的交互式多区域模型(UT-IMRM)算法.UT-IMRM对每个子模型使用无迹卡尔曼滤波器(UKF)进行滤波.在目标跟踪实验中对UT-IMRM性能进行测试.实验结果显示当系统参数不属于IMM模型集合时,UT-IMRM能够比IMM获得更好的估计性能.

关键词: 交互式多模型, 参数估计, 无迹卡尔曼滤波器

Abstract: A nonlinear filtering method called Interacting Multiple Region Model (IMRM) is proposed to estimate the state and continuous system parameter together.IMRM regards the continuous system parameter space as a set of disjoint sub-regions,and each sub-region is assigned to a sub-model respectively.IMRM runs a bank of sub-filters in parallel.At each time step,IMRM computes the mixed initial condition for each sub-model by interaction operation,and each sub-filter estimates the state and system parameter on the condition that the system parameter belongs to a unique sub-region.In order to implement the IMRM efficiently,Unscented Transformation based IMRM (UT-IMRM) is developed by using the unscented kalman filter as the sub-filter.A target tracking experiment is designed to test the performance of UT-IMRM.Experimental results show that UT-IMRM achieves better estimation performance than that of IMM when the system parameter doesn’t belong to the IMM model set.

Key words: interacting multiple model, parameter estimation, unscented Kalman filter

中图分类号: