Adaptive Alignment for Low-Cost INS in ECEF Frame Under Large Initial Attitude Errors

A new two-stage alignment on a moving base is proposed for a low-cost inertial navigation system with the aid of a Global Positioning System (GPS) receiver in the Earth-centered, Earth-fixed (ECEF) frame under large initial attitude errors. Both the state and measurement equations are derived using the additive quaternion error model in the ECEF frame for alignment without the need for level adjustment. The nonlinear velocity error equations in both the coarse and the fine stages are linearized further. An improved strong tracking filter is proposed for the coarse stage to converge quickly with large initial attitude errors. After the attitude errors converge to smaller values in the coarse stage, a standard Kalman filter is implemented for the fine stage. Experiments demonstrate that attitude error can converge quickly with the proposed adaptive alignment method.