Improvement of UKF Algorithm and Robustness Study

Abstract

Iterated unscented Kalman filter (IUKF) algorithm has improved the unscented Kalman filter (UKF) and enhanced the performance of filter estimation by using Newton-Raphson iterative equation. This paper improves IUKF algorithm ulteriorly after detailedly analyzing principle of IUKF and its iterative equation, and proposes a new filtering algorithm with robustness-Improved IUKF. Then the performance of the new algorithm is validated by two experiments. The results show that the improved IUKF is more robust which can effectively resist the influence of measurement outlier.