Target tracking with regularized variable rate particle filters

Recently introduced variable rate particle filters (VRPF), utilizing semi-Markov models for maneuvering target tracking obtained superior performance compared to the standard Markov models [1]. However, degeneracy problem commonly encountered in particle filtering also arises in VRPF algorithm. In this work, regularization technique used in standard particle filtering [2] is integrated to VRPF modelling and regularized variable rate particle filters (R-VRPF) are introduced as a solution to degeneracy problem. Performance of proposed R-VRPF algorithm is investigated in a bearing only target tracking problem and it is shown that RMS position error is reduced due to the better approximation to the posterior distribution representing target position.