A mixed-type accurate continuous-discrete extended-unscented kalman filter for target tracking

This paper presents a novel method of nonlinear Kalman filtering, which unites the best features of the accurate continuous-discrete extended Kalman and unscented Kalman filters. More precisely, the time updates in the discussed state estimator are done by the corresponding part of the first filter whereas the measurement updates are conducted with use of the unscented transformation. All this allows accurate predictions of the state mean and error covariance to be combined with accurate measurement updates. Therefore the new filter is particularly effective for stochastic continuous-discrete systems with nonlinear and/or nondifferentiable observations. The efficiency of this mixed-type filter is shown in comparison to the performance of the accurate continuous-discrete extended Kalman and unscented Kalman filters on a known target tracking problem with sufficiently long sampling periods.