Numerical implementation of continuous-discrete IMM state estimators

This paper deals with the problem of numerical implementation of adaptive tracking filters based on continuous-time dynamic models of target motion and on discrete-time models of radar measurement process. The particular problem considered is characterised by: non-linear and non-stationary dynamic models of target movement, uncertain parameters and relatively low data rate due to a rotating antenna. The proposed tracking filter relies basically on the continuous-discrete variant of the extended Kalman filter (EKF), the probabilistic data association (PDA) technique and the interacting multiple-model (IMM) state estimation scheme.