Adaptive In-Flight Alignment Method for SINS/GPS/Polarization Integrated Navigation System Based on the Degree of Observability Analysis

In-flight alignment is one of the key technologies of inertial navigation systems (INSs). The alignment process requires not only high accuracy and rapidity but also robustness to unknown noise and external environment. For the strapdown INS (SINS)/GPS/polarization integrated navigation system, an adaptive in-flight alignment method based on the degree of observability analysis (DOA) is proposed in this article. First, polarization navigation is used to improve the attitude accuracy of the SINS/GPS integrated navigation system. Second, an adaptive federated Kalman filter (AFKF) is applied to fuse navigation information. In the local filter, the variance matrix of measurement noise is adjusted by the innovation of the measurement value. In the master filter, the degree of observability (DoO) of the state variable is used to adaptively determine the information allocation factor (IAF) for information fusion. The proposed adaptive in-flight alignment method can adaptively adjust the noise variance matrix according to the external environment, and automatically update the IAF in the AFKF with various maneuver modes. Finally, compared with the general federated Kalman filtering method, the experimental result shows that the adaptive federated filtering method proposed in this article not only improves the alignment accuracy of the navigation system but also improves the robustness.