Tracking algorithm with radar and infrared sensors using a novel adaptive grid interacting multiple model

This study presents a novel adaptive grid interacting multiple model based on modified iterated extended Kalman filter (AGIMM-MIEKF) for tracking a manoeuvreing target using radar/infrared (IR) heterogeneous sensors. This tracking algorithm is developed by aligning observation data of radar/IR sensors in time, and fusing the synthesised data before applying to AGIMM-MIEKF algorithm. Under the architecture of the proposed algorithm, the AGIMM deals with the model switching, whereas the MIEKF accounts for non-linearity in the dynamic system models. A new measurement update equation and an iterated termination criterion are derived and applied to radar/IR tracking system. The simulation results show that the presented AGIMM-MIEKF has higher tracking precision than the traditional algorithms.