Iterated square root unscented Kalman particle filter

Abstract

In order to improve tracking estimation accuracy of square-root unscented Kalman particle filter (SRUKFPF), a new particle filter algorithm of update SRUKF based on iterated measurements is proposed. The algorithm produces the important density function of particle filter using maximum posteriori estimate of iterated square-root unscented Kalman filter, and amends the state covariance using Levenberg-Marquardt method, so that the observed information of particle is effectively used. This is more consistent with the posterior probability distribution of true state. Simulation results show that estimation performance of the proposed algorithm is much better than standard particle filter (PF), unscented particle filter (UPF) and square root unscented Kalman particle filter (SRUKFPF).