Abstract: In maneuvering target tracking，the movement model is non-linear or the noise is uncertain and so on.To solve these problems，an Interacting Multiple Model-Unscented Particle Filter（IMM-UPF） is designed by combining Interacting Multiple Model（IMM），Particle Filter（PF） and Unscented Kalman Filter（UKF）.In this algorithm，the multiple models structure is used to track arbitrary maneuvering of the target，particle filter is used in every model to deal with the nonlinear and non-Gaussian problems，and UKF is used to produce particles，thus，the distribution of particles is closer to the posterior probability density distribution and the degradation of particle phenomenon is overcome，which enhances the accuracy of estimations for the current observations accounted.The model set of the system contains three models based on the actual movement model.At the end，the results demonstrate the efficiency of the new filtering method through the example of simulation，and the performance of IMM-UPF is better than PF，UPF algorithms.

Key words: target tracking, interacting multiple model, Particle Filter（PF）, Interacting Multiple Model-Unscented Particle Filter（IMM-UPF）

摘要： 为解决机动目标跟踪的非线性和噪声不确定等问题，提出了一种新的滤波算法：融合了交互式多模型（IMM）、粒子滤波（PF）和无迹卡尔曼滤波（UKF）的IMM-UPF算法。该算法采用多模型结构以跟踪目标的任意机动，粒子滤波能处理非线性、非高斯问题，而采用UKF产生粒子，由于考虑了当前观测值，使得粒子的分布更接近后验概率密度分布，克服粒子的退化现象，从而提高估计精度。系统的模型集根据实际的目标系统设计了三个非线性模型。通过实例仿真，结果证明了IMM-UPF算法的有效性，且其性能优于PF、UPF算法。

关键词: 目标跟踪, 交互式多模型, 粒子滤波, 交互式多模型无迹卡尔曼滤波（IMM-UPF）