Improved robust Huber-based divided difference filtering

Abstract

This article derives an improved robust Huber-based divided difference filter by using the Huber’s technique, in which the nonlinear measurement function is directly used in the nonlinear regression equation instead of the linear or statistical approximation. The presented filtering algorithm exhibits robustness against the deviations from the Gaussian error distribution and has better estimate accuracy compared with the Huber-based divided difference filter. This filter is applied to a benchmark problem of estimating the trajectory of an entry body from discrete-time range data measured by a radar tracking station. Simulation results indicate that the proposed filter algorithm outperforms the previous methods in terms of robustness and accuracy.