A Robust Cubature Kalman Filter for INS/CNS Integrated Navigation with Measurement Outliers

To tackle the reduced filtering efficacy in INS/CNS integrated navigation systems faced with measurement outliers, this study introduces a robust cubature Kalman filter (RCKF) that employs hypothesis testing to boost the CKF's resilience to these outliers. Within the RCKF framework, a chi-square test (HT) along with an enhanced sequential probability ratio test (SPRT) are employed for online estimation and adjustment of the measurement noise covariance. Simulation experiments confirm the effectiveness of the proposed approach.