Abstract:

An on-line updating method of Markov transition probability for the interacting multiple model(IMM)algorithm is proposed, and the square-root cubature Kalman filter(SRCKF)is introduced into IMM, so a novel time-varying Markov transition IMM-SRCKF algorithm is obtained.Using real-time recursive estimation method based on the system mode information implicit in the current measurements, the proposed algorithm effectively avoids the problem of prior determination of the Markov transition probability matrix in traditional IMM. Furthermore, SRCKF propagates the square root of the covariance in filter interaction so that it guarantees the symmetry and positive semi-definiteness of the covariance matrix and greatly improves the numerical stability and numerical accuracy. Simulation results show that the proposed algorithm has better tracking performance and higher efficiency compared with the conventional IMM and IMMCKF.