Pearson type VII distribution‐based robust Kalman filter under outliers interference
Yun Peng, Panlong Wu, Shan He · IET Radar Sonar & Navigation · 2019
In this study, the authors focus on state estimation in a linear discrete time system where measurements are uncertain due to possible outliers. A novel robust Kalman filter based on the Pearson type VII (PTV) distribution is proposed by using the variational Bayesian (VB) method and tested in a target tracking example. To judge whether the measurement is an outlier, a judgement factor that follows the beta‐Bernoulli distribution is introduced in the measurement modelling. In addition, the outlier noise is approximated by the PTV distribution considering the heavy tail characteristic and it is jointly estimated with the judgement factor and state using the VB method. The simulation results show that the proposed filter has better estimation accuracy than most existing robust filters.