Title :
Novel cubature Kalman filters based on mixed degrees
Author :
Yali Feng;Qingan Yang;Shiyuan Wang;Wenjie Zhang;Yunfei Zheng
Author_Institution :
School of Electronic and Information Engineering, Southwest University, Chongqing, China
Abstract :
The cubature rule used in the cubature Kalman filter (CKF) is a combination of the spherical rule and the radial rule. In order to further improve the computational efficiency and accuracy, we introduce a novel CKF in this paper by using the mixed degrees of spherical and radial rules. Simulation results show that the novel CKFs present the trade-off between the conventional third-degree CKF (CKF3) and the fifth-degree CKF (CKF5) from the respects of the accuracy and computational complexity. In addition, the novel CKF outperforms the CKF5 in some nonlinear systems.
Keywords :
"Kalman filters","Bayes methods","Mathematical model","Computational complexity","Time measurement","Noise measurement"
Conference_Titel :
Intelligent Control and Information Processing (ICICIP), 2015 Sixth International Conference on
Print_ISBN :
978-1-4799-1715-0
DOI :
10.1109/ICICIP.2015.7388172