A novel adaptive filter for highly maneuvering target

This paper considers the highly maneuvering target tracking as a discrete-time stochastic system with generalized unknown disturbance input, which can represent an arbitrary linear combination of dynamic, random, and deterministic disturbance inputs. The proposed filter based on upper bound filter (UBF) can adaptively optimize adjust factor to find the globally optimal solution of covariance matrices of the state predictions, innovation and estimates. To reduce the linearization error, a iterative optimization frame is designed. The experiment on 'Smaneuver' of anti-ship missile shows that the proposed filter can significantly reduce the peak estimation errors due to maneuvers.