自适应模型集交互式多模型算法

针对交互式多模型(IMM)算法的目标跟踪精度问题,提出了一种自适应模型集IMM算法.利用IMM算法中的模型概率含义,并以此对模型集的收缩比例因子进行设计,这样模型集通过向中心模型收敛可完成自适应调整,而自适应调整过程能有效、实时地利用观测信息.仿真实验结果表明,所提算法能有效跟踪机动目标,而且比IMM算法的跟踪精度更高,但其受到目标机动模型的先验性的限制.

Multiple Model Algorithm Based on Adaptive Interaction of Model Set

In view of the track performance of the interacting multiple model(IMM) algorithm in the target tracking,an adaptive IMM algorithm of model set is presented.The implication of the model probability in the IMM algorithm is utilized to design the stretching factor of the model set.Thus,the model set can be adjusted adaptively by converging it to the center model,in which the real-time observations are utilized efficiently.Simulation results show that the maneuvering target is effectively tracked by using the proposed algorithm and the track accuracy is higher compared to conventional IMM algorithm.The limitation of the algorithm is that the prior information about the target maneuvering model must be known.