基于模型调整的自适应交互多模型算法

An adaptive interacting multiple model algorithm based on model adjustment

针对交互多模型算法的技术特点,利用量测中所包含的当前信息对目标模型集的自适应调整并对调整后的模型概率进行估计,实现了模型集的自适应交互多模型算法。介绍了这种算法的设计步骤和仿真方法。仿真结果表明了该算法比标准IMM算法具有更高精度的跟踪性能。

On the basis of the characteristics of Interacting Multiple Model（IMM） used in target tracking, we put forward an adaptive IMM based on model adjustment. The current information contained in the measurement is used for adaptive adjustment of target model set, and estimation is made for time-varying Markov transition probabilities of the model. Thus the model set and transition probability can be changed properly during tracking, and the adaptive IMM of model set is realized. The design process and simulation method of the algorithm is introduced here. Simulation result showed that： compared with standard IMM, the adaptive IMM is more effective and has higher precision in tracking.