IMM-Singer模型的机动目标跟踪算法

摘要

The Interacting Multiple Model (IMM) algorithm is an effective solution to maneuvering target tracking, however, its performance depends on the type, number and parameters of the models. The Singer Model algorithm can be utilized for maneuvering target tracking except that the parameters such as maneuver frequency, process noise and so on are hardly set. To the problems as mentioned above, with the merit that IMM can be combined with other algorithms easily, a novel filter of IMM-Singer Model is proposed such that the parameters of Singer Model can be set automatically. Simulation results show that compared with the single Singer Model and the general IMM, the proposed algorithm is more effective in improving the accuracy of maneuvering target tracking.%交互式多模型(IMM)算法是一种有效机动目标跟踪算法,但其性能与模型的选择、个数以及参数有关.Singer模型算法可以实现对机动目标的跟踪,但该算法存在机动频率和过程噪声大小等参数难以选取的问题.针对以上情况,利用IMM算法易于结合其他算法的特点,提出一种基于IMM-Singer模型的机动目标跟踪算法,实现Singer模型参数的自适应选择.仿真结果表明,该算法比单一的Singer模型算法或一般的IMM算法更能有效提高机动目标跟踪精度.