Title :
Relations Between Sparse-Grid Quadrature Rule and Spherical-Radial Cubature Rule in Nonlinear Gaussian Estimation
Author :
Bin Jia ; Ming Xin ; Yang Cheng
Author_Institution :
Intell. FusionTechnology, Germantown, MD, USA
Abstract :
The recently emerging cubature Kalman filter and the sparse-grid quadrature filter approximate the numerical integrations for the mean and covariance in Gaussian filters using the spherical-radial cubature rule and the sparse-grid quadrature rule, respectively. This technical note reveals that 1) spherical rules can be obtained by the projection of sparse-grid quadrature rules; 2) the third- and some of the fifth-degree spherical-radial cubature rules can be directly constructed from sparse-grid quadrature rules.
Keywords :
Gaussian processes; Kalman filters; nonlinear estimation; nonlinear filters; numerical analysis; Gaussian filters; cubature Kalman filter; fifth-degree spherical-radial cubature rules; nonlinear Gaussian estimation; numerical integrations; sparse-grid quadrature filter; sparse-grid quadrature rule; Accuracy; Approximation methods; Estimation; Indexes; Kalman filters; Polynomials; Kalman filter; nonlinear filtering; sparse-grid; spherical-radial cubature rule;
Journal_Title :
Automatic Control, IEEE Transactions on
DOI :
10.1109/TAC.2014.2322478