A redundant adaptive robust filtering algorithm based on cubature Kalman fliter

In this paper, A novel nonlinear state estimation algorithm called redundant adaptive robust filter (RARCKF) is proposed for the state estimation of the maneuvering target in which the square-root of the cubature kalman filter (SRCKF), like the other traditional Gaussian domain Bayesian filters, cannot achieve high accuracy of state estimation when it suffers from long-standing model errors or the model of the system takes rapid and abrupt unknown changes. As a result of using RARCKF, the algorithm can make sure the validity of the filter while in the case of the model prediction suffers with long-standing errors or the target takes maneuvering. Simulation results in the section 4 indicate RARCKF outperforms over the SRCKF both in the numerical accuracy and the convergence rate.