A Novel Gaussian Nonlinear Iterated Update Filtering

To solve the problem of performance degradation and divergence of the Gaussian nonlinear filter in the large initial deviation condition, a new nonlinear filtering method is proposed, which is the Iterated Update Extended Kalman Filter (IU-EKF). The new approach is carried out by introducing the current time measurement information gradually to the measurement update process in pseudo-time. The proposed algorithm has two advantages, one is that it increases the filtering accuracy substantially, compared with the Extended Kalman Filter, the other is that it is superior to the classical Gaussian hypothesis filter in large prior uncertainty condition. Simulation has been taken to compare the filtering performance among the EKF, IEKF, UKF, CKF and the proposed method. The simulation results demonstrate higher accuracy provided by the new filter than the classical Gaussian approximate filter mentioned above.