基于UKF的马尔可夫参数自适应IFIMM算法

Innovation Filter Interacting Multiple Model with Adaptive Markov Parameter Based on UKF Algorithm

给出了一种基于不敏卡尔曼滤波(UKF)的马尔可夫参数自适应的新息滤波器交互式多模型算法,较好地解决了非线性条件下机动目标跟踪的问题,可获得比基于扩展卡尔曼滤波的交互式多模型(IMM)算法和基于UKF的IMM算法更好的稳定性和计算精度,还避免了复杂的Jacobi矩阵运算;该算法结合了马尔可夫参数自适应和新息滤波器技术,实现了马尔可夫转移矩阵的自适应和量测噪声的减小。最后,通过Monte Carlo仿真进一步验证了该方法的正确性和有效性。

Innovation Filter Interacting Multiple Model（IFIMM） with adaptive Markov parameter based on UKF algorithm is pres- ented in this paper, which can effectively deal with maneuvering target tracking problems in practice. This algorithm exhibits better computational stability and precision than those based on EKF and UKF, and avoids computing complicated Jacobi matrix. Moreo- ver, using adaptive Markov parameter and innovation filter techniques, this algorithm can modify Markov transition matrix in real time and filter part of the measurement noise. With numerous Monte Carlo simulations, the efficiency and effectiveness of the algo- rithm are shown.