A fault-tolerant sins/dual 2D-LDV tightly coupled integration scheme for autonomous vehicle navigation

Strapdown inertial navigation systems (SINS) integrated with two-dimensional laser Doppler velocimeters (2D-LDVs) present a promising autonomous navigation solution for land vehicles, particularly in GNSS-denied environments. However, their performance is often degraded by vehicle sideslip and outliers in 2D-LDV measurements. This paper addresses these challenges by proposing a novel fault-tolerant SINS/Dual-2D-LDV tightly coupled integration scheme. In this scheme, two 2D-LDVs are integrated with SINS to create a redundant measurement model. This model utilizes the raw measurements from both LDVs along with the vehicle's lateral zero-velocity constraint. To handle anomalies, a fault detection method based on the Local Outlier Factor (LOF) is introduced to identify measurement outliers and violations of the zero-velocity constraint. An adaptive filter, whose gain is dynamically adjusted by the LOF value, is then employed to mitigate the impact of these anomalies on the integrated navigation solution. The effectiveness and robustness of the proposed method are validated through two sets of long-distance vehicle experiments. Results demonstrate that the proposed scheme achieves superior positioning accuracy in both horizontal and vertical directions compared to traditional approaches. Furthermore, the LOF-based fault detection method proves to be more sensitive and effective in identifying anomalies than the traditional residual chi-squared detection method, enhancing the overall reliability of the system.