用于非线性机动目标跟踪的新型IMM算法1

  文章编号: 1671- 637  ( 2008) 08001406

用于非线性机动目标跟踪的新型 IMM 算法

孙庆鹏,  孔祥维,  卢聪聪,  邓江安

( 大连理工大学信息学院, 辽宁 大连 116024)

摘   要:  针对在非线性机动目标跟踪中存在的滤波器易发散、机动检测有延迟等问题, 把

Unscented Kalman Filter(UKF) 引进到交互多模型算法( IMM) 中, 设计了交互多模型 UKF 滤波器。并

利用目标运动模型集概率的相对变化率设计了自适应交互多模型UKF 滤波器, 最后进行了计算机

仿真。蒙特卡罗仿真结果表明, 两种滤波算法都具备UKF 滤波器精度高、稳定性好、不易发散的优

点, 同时不需了解目标机动的先验信息, 适合于实际应用; 并且自适应交互多模型UKF 滤波器具有

Abstract:  In nonlinear maneuvering target tracking, the tracking filters are liable to diverge or have detecting

delays. To solve the problem, an Interacting Multiple Model( IMM) U nscented Kalman Filter ( UKF) is de

signed by introducing UKF into IMM algorithm. And a new adapt ive interacting multiple model UKF algorithm

is also designed by using the relative variance ratio of the probability of target motion model set. The Monte

Carlo simulation results indicate that both the two proposed IMM filters do not rely on a prior knowledge about

the target mot ion, and have the advantages of high accuracy and good stability. At the same time, they almost do

not diverge, which will be effective in realtime target tracking. The adaptive interacting multiple model UKF

presents better tracking performance.

Key words:  maneuvering target tracking;  adaptive filter;  interact ing multiple model algorithm

第 15 卷第 8 期

2008 年 8 月

电光与控制

Electronics Optics & Control

Aug. 2008