Adaptive filter design for low-cost MEMS-IMU/GPS integrated navigation

It is essential to consider the affect of circumstance and stability of apparatus when designing a practical filter for low-cost MEMS-IMU/GPS integrated navigation. Innovation adaptive estimation (IAE) can adjust noise variance covariance online. On the other hand, IAE is more sensitive to observation outliers. Based on the orthogonal property of innovation sequence, the paper present the improved adaptive filter which inspect innovation orthogonal property and assign weight to innovation with outliers. The algorithm is embedded into MEMS-IMU/GPS tightly integrated navigation filter. Experiment shows that modified algorithm can restrain observation outliers efficiently, that improve the robustness of integrated system.