An adaptive mode transition probability IMM Bernoulli filter
Feng Ping Yang, Wanying Zhang, Yazhe Su, Yao Xuanzheng · 2016
The Bernoulli filter (BF) is a recursive target tracking mechanism based on the random finite set (RFS) theory. For the tracking of maneuvering target, the changes of target motion model can be solved by using interacting multiple model (IMM) estimator. However, in practical systems, mode transition probabilities may be related to the state and varies with time, it is unreasonable to assume constant transition probabilities. This paper introduce prior road information to establish an adaptive mode transition probability, and combine with IMMBF to achieve ground maneuvering target tracking. Simulation results show that the proposed algorithm outperforms the constant mode transition probability algorithm.