An IMMUPF method for ground target tracking

In this study, a new interacting multiple model unscented particle filter (IMMUPF) for tracking ground maneuvering targets is presented. A bank of unscented particle filters is used in the interacting multiple model (IMM) framework for updating the state of moving target. Two groups of multiple model filters: IMM-based filters and particle-based multiple model filters, are compared for their capability in dealing with ground maneuvering target tracking problem. Simulation results show that particle-based filters outperform IMM-based filters in the estimate accuracy and the IMMUPF method relatively has much better performance than the IMMPF method.