用于非线性机动目标跟踪的新型IMM算法

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滤波器具有更好的跟踪效果。

Two new IMM algorithms for nonlinear maneuvering target tracking

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) Unscented Kalman Filter(UKF) is designed by introducing UKF into IMM algorithm.And a new adaptive 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 motion,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.