The augmented form of cubature Kalman filter and quadrature Kalman filter for additive noise

In this paper, the augmented forms of the quadrature Kalman filter (QKF) and cubature Kalman filter (CKF) are presented for estimating the nonlinear dynamic systems. The QKF and CKF are modified by forming an augmented state variable, which concatenates the state and noise components together, so that the effect of process and measurement noises can be used to better capture the odd-order moment information. The simulation results demonstrate the improved performance of the augmented form over the nonaugmented form. Besides, the performance and the execution time of the three kinds of nonlinear filters is also compared in the augmented form and nonaugmented form.