Robust wavelet-based inertial sensor error mitigation for tightly coupled GPS/BDS/INS integration during signal outages

This paper proposes a robust wavelet-based tightly coupled Global Positioning System (GPS)/ Beidou Navigation Satellite System (BDS)/Inertial Navigation System (INS) integration scheme aiming to improve the overall position accuracy during signal outages. A robust wavelet denoising model based on α-trimmed mean filter demonstrates its effectiveness on noise reduction and gross error elimination of inertial sensor raw data. Thereafter, a robust wavelet-based tightly coupled GPS/BDS/INS integration scheme is proposed, and GPS/BDS double-difference (DD) carrier-phase and pseudorange measurements are introduced to build a 27-state tightly coupled GPS/BDS/INS integration equation. The extended Kalman filter (EKF) has been designed for state estimation, and the inclusion of BDS enhances the satellite geometric strength of the whole navigation system. A field vehicle test indicates the position accuracy of five simulated GPS/BDS outages can be improved by about 7, 16, and 33% for north, east, and up components with the proposed scheme compared to standard integration scheme, and gross errors have been detected and eliminated with the new integration scheme. The authors also find that the BDS phase residual is larger than GPS and satellite dependent in the tightly coupled GPS/BDS/INS integrated navigation system.