IEEE Sensors Journal · 2021 · 62 citations · 31 references
ComputingComputational EngineeringGeodesyGlobal Navigation Satellite SystemsUncertainty QuantificationBayesian FilteringKalman FilteringParameter Estimation Under UncertaintyState EstimationAdaptive Kalman FilterBayesian Variational InferenceCovariance EstimationNonlinear FilteringNumerical SimulationRobust Kalman FilterVariational Bayesian FilteringVariational Bayesian Inference
In this article, we focus on addressing the nonlinear filtering problem with unknown measurement noise covariance and measurement outliers, which may be encountered in the application in strapdown inertial navigation system/global positioning system integrated navigation system. Although the existing methods, such as the adaptive Kalman filter, are widely used in the integrated navigation system, their estimation accuracy is poor, this paper proposes a variational Bayesian-based robust cubature Kalman filter to address this problem, which not only retains the adaptivity when the measurement noise is unknown but also exhibits robustness in the presence of measurement outliers, First, the variational Bayesian method is applied to the estimation of measurement noise, then the maximum correntropy criterion is introduced to replace the minimum mean square error criterion, which improves the robust performance of the filter. The numerical simulation demonstrates that the proposed filter outperforms the existing filters both in estimation accuracy and robustness, and the effectiveness of the proposed filter is verified on the integrated navigation system.
31
Ienkaran Arasaratnam, S. Haykin · IEEE Transactions on Automatic Control · 2009 · 3.3K citations
Computing, Computational Engineering, Bayesian Data Analysis +11
Adaptive Kalman Filtering for INS/GPS
Ahmed Mohamed, K. P. Schwarz · Journal of Geodesy · 1999 · 1.1K citations
Computing, Geophysics, Geodesy +9
Maximum correntropy Kalman filter
Badong Chen, Xi Liu, Haiquan Zhao et al. · Automatica · 2016 · 868 citations
J.H. Kotecha, Petar M. Djurić · IEEE Transactions on Signal Processing · 2003 · 806 citations