High-degree cubature Kalman filter for nonlinear state estimation with missing measurements

This paper proposes high-degree cubature Kalman filter for nonlinear systems with missing measurements. We derive out the explicit formulas for the prediction and update in the filtering. To fulfill the numerical computation, especially the numerical integrals, of these formulas, the fifth-degree spherical-radial cubature rule is adopted to give a high-degree cubature Kalman filtering algorithm. Through numerical example, it is shown that the fifth-degree cubature Kalman filter has better precision and stability than the extended Kalman filter, the unscented Kalman filter, and the fifth-degree unscented Kalman filter.