• Proposes a novel GPS/INS integration approach based on the Covariance Matching Adaptive Kalman Filter (CMAKF). • Dynamically adjusts covariance matrices using real-time statistical consistency to enhance adaptability. • Addresses performance degradation of conventional Kalman Filters in dynamic and unpredictable environments. • Achieves significant positioning error reduction (81% vs. KF, 25% vs. FKF, and 19% vs. IF). • Demonstrates superior robustness and stability under diverse and challenging land-based navigation scenarios. The integration of the Global Positioning System (GPS) and Inertial Navigation System (INS) harnesses the complementary advantages of both systems to achieve precise and dependable positioning. Standalone INS solutions accumulate errors over time, necessitating augmentation. Traditionally, the Kalman Filter (KF) has been employed for GPS/INS integration, delivering strong performance under controlled conditions. However, their effectiveness diminishes in dynamic and unpredictable environments, particularly during extended operation. To address these limitations, this study introduces a novel GPS/INS integration method based on the Covariance Matching Adaptive Kalman Filter (CMAKF). This approach dynamically adjusts the filter's covariance matrices based on real-time statistical consistency, improving the system's adaptability to environmental changes and sensor errors. The proposed method yields substantial improvements in GPS/INS integration performance compared with the standard KF, achieving an average positioning error reduction of 81% relative to KF and 25% and 19% relative to Faded Memory KF (FKF) and Information Filter (IF), respectively. Moreover, empirical stability assessments demonstrate that CMAKF exhibits superior consistency and robustness, particularly under diverse and challenging data conditions, establishing it as a highly effective strategy for advanced land-based positioning applications.
Alaeiyan et al. (Sun,) studied this question.