Abstract:

According to the limitation that the conventional cubature Kalman filter (CKF) requires system and measurement noise to be uncorrelated, a novel CKF with correlative noises for nonlinear discrete time Gaussian systems is designed. A set of recursive filtering equations of CKF with correlative noises are derived based on Bayesian estimation rule, and the third-order spherical-radial cubature rule is utilized to approximate the postrior mean and covariance of the state. The proposed method can estimate the state as a conventional CKF is unavailable when the system and measurement noise are correlative Gaussian white noises, which expends the application of CKF. The effectiveness of the proposed method is verified by a numerical simulation example.