Sequential Integrity Monitoring With State-Domain Consistency Detection for Integrated Navigation in Adverse Urban Areas

Integrity monitoring (IM) quantitatively assesses the confidence of position solutions, playing a crucial role in safety-critical autonomous applications. This study proposes a sequential multiple faults-based IM method for the global satellite and inertial integrated navigation system (GINS) in challenging urban environments. Initially, a consistency factor in the state domain is calculated using the sequential probability ratio over sliding windows, which reflects the characterizing effect of the filter-indicated state mean square error (mse) on the position error distribution. Subsequently, the potential position bias is derived through the filtering innovation-based maximum eigenvalue under the multifault missed assumption. A protection level is then determined by the filter-indicated MSE and the position bias induced by potential multiple errors combined with the consistency factor, which overbounds the position error accurately with reduced redundancy space, effectively assessing the confidence of the GINS position solution. Finally, an in-vehicle experiment demonstrates that the proposed method has the highest protection level reliability of 99.15% throughout the experiment. Moreover, the probability of hazardous misleading information is more than 7.76% lower than that of existing methods in severe adverse areas. Thus, the effectiveness of the proposed method has been verified.