An Innovative Internal Model Adaptive Kalman Filtering Method and Its Applications

To deal with the stochastic signal processing problem in which the deterministic disturbance is involved, a new adaptive Kalman filtering technique based on internal model (so-called Internal model adaptive kalman filtering──IMAKF) is proposed. The basic idea is as follows: First, the internal models of both the signal and the deterministic disturbance in the observed data are established by means of piecewise sine wave─fitting. Furthermore, the parameters in these internal models are taken as the augmented state variables to form a new nonlinear system model. Then, the iterative version of extended Kalman filtering (EKF) algorithm is utilized to realize the real-time tracking of those internal model parameters. The innovative approach has been successfully applied to the GPS signal estimation for the purpose of maneuvering target tracking. In comparison with the existing methods, the proposed approach can remarkably improve the positioning accuracy.