Robust adaptive extended target filtering based on multiplicative noise model
Li Wang, Hui Chen, Feng Lian, Xuxin Wang · Physica Scripta · 2025
Abstract This study proposes two novel methods for tracking extended target with elliptical shape under non-Gaussian noise conditions. To achieve this, an explicit nonlinear measurement equation is formulated by incorporating a multiplicative noise term that connects the target's kinematic and shape parameters with the measurements. This measurement equation is employed to derive a closed-form solution for the Student's t extended Kalman filter based on multiplicative noise model (St-EKF-MNM) with recursive measurement updates. To further optimize the target tracking performance, the Hessian matrix was used to deal with the nonlinearity in the measurement equation, further derive the Student's t second-order extended Kalman filter based on multiplicative noise model (St-SOEKF-MNM). The effectiveness and robustness of the proposed methods are demonstrated with simulation.