An Effective Outlier Suppression Technique for Moving Horizon Estimators

This paper investigates the state estimation problem in discrete-time linear systems affected by measurement outliers. The conventional moving horizon estimation (MHE) methods suffer from poor estimation accuracy and low computational efficiency. To address these issues, a novel outlier-robust MHE strategy has been designed in this study. Specifically, at each sampling instant, a set of least-squares cost functions is provided to detect outliers and provide state estimates. The stability of the proposed estimator is demonstrated. The estimation error can achieve exponential convergence by choosing appropriate design parameters. Simulation results demonstrate that the proposed estimation strategy outperforms the existing state-of-the-art MHE methods with outlier suppression.