A Gaussian Uniform Mixture Model for Robust Kalman Filtering

This paper presents a Kalman-type recursive estimator for discrete time systems with a measurement noise modeled by a Gaussian-uniform mixture. The objective is to deal with data containing outliers that degrade the performances of the regular Kalman filter. The introduced non-Gaussian noise model takes into account the reliability of the measurement to be robust with respect to erroneous data. The Kalman-type estimator is based on the Masreliez's formulation that copes with non-Gaussian noise models. Results in different simulated conditions are displayed to evaluate the performances of the introduced algorithm and to compare it to state-of-art alternatives.