Enhanced Fault Detection and Exclusion in GNSS/INS Integrated Navigation Systems via Robust Estimation

A robust estimation-based fault detection and exclusion algorithm is proposed for GNSS/INS integrated navigation systems to mitigate the error tracking effect that limits traditional innovation monitoring methods. By adaptively adjusting the Kalman filter gain using the IGG-III weight function within a sliding window framework, the proposed approach enhances sensitivity to ramp faults and improves fault handling robustness. Simulation results demonstrate up to 34% faster detection in ramp fault scenarios and the successful isolation and exclusion of faults that remain undetected by traditional methods, underscoring its effectiveness in improving reliability and robustness in integrated navigation systems.