一种新的基于交互多模型的序贯重要采样算法

A Novel Sequential Importance Sampling Algorithm Based on Interacting Multiple Model

通过将交互多模型(IMM)算法和粒子滤波(SIS)算法结合,提出了一种新的IMM-SIS算法。在每个模型中,都有一个标准的粒子滤波器,模型之间的交互与传统的IMM一样。由于在新的算法中,每个模型中粒子滤波都保证固定数量的粒子,因此不会出现粒子退化和贫乏现象。仿真证明了新的IMM-SIS算法在收敛速度和精度方面都要优于传统的IMM-EKF算法。

A novel method for multiple model particle filtering for Markovian switching systems is introduced. This new method is a combination of the interacting multiple model （IMM） filter and particle filter. A regularized particle filter is running in every mode. The mixing and interaction is similar to that in a conventional IMM filter. Furthermore, the new method keeps a fixed number of particles in each mode, and therefore it avoids the particle degeneracy and impoverishment phenomenon. Simulations show that the tracking speed and accuracy of the IMM - SIS algorithm are better than those of the IMM - EKF.