A new interacting multiple model algorithm for maneuvering target tracking based on adaptive transition probability updating

Abstract

A new interacting multiple model (IMM) algorithm for maneuver target tracking is proposed in this paper. An adaptive updating method of the model transition probability is developed based on Bayes Theorem. Reinitialization strategies of the transition probabilities are design for two special cases. Then the updated transition probabilities are used to fulfill the interaction of the models. The proposed algorithm is of lower computational burden. Numerical simulations for both filtering and smoothing are given to demonstrate estimation performance.