An Adaptive Unscented Kalman Ilter Integrated Navigation Method Based on the Maximum Versoria Criterion for INS/GNSS Systems

• Published in: Sensors (Basel, Switzerland) Vol. 25; no. 11; p. 3483
• Main Authors: Zhang, Jiahao, Feng, Kaiqiang, Li, Jie, Zhang, Chunxing, Wei, Xiaokai
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 31.05.2025; MDPI
• Subjects: Accuracy; Algorithms; Analysis; Artificial satellites; Electronics in navigation; Inertial navigation; Inertial navigation (Aeronautics); INS/GNSS-integrated navigation; Kalman filters; maximum versoria criterion; Methods; Navigation systems; non-Gaussian noise; Optimization; Sensors; unscented Kalman filter; INS/GNSS-integrated navigation; maximum versoria criterion; non-Gaussian noise; unscented Kalman filter
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s25113483

Aimed at the problem of navigation performance degradation in inertial navigation system/global navigation satellite system (INS/GNSS)-integrated navigation systems due to measurement anomalies and non-Gaussian measurement noise in complex navigation environments, an adaptive unscented Kalman filter (AUKF) algorithm based on the maximum versoria criterion (MVC) is developed. The proposed method is designed to enhance INS/GNSS-integrated navigation system robustness and accuracy by addressing the limitations of conventional filtering approaches. An adaptive unscented Kalman filter is constructed to enable dynamic adjustment of filter parameters, allowing for real-time adaptation to measurement anomalies. This ensures accurate tracking of navigation parameter states, thereby improving the robustness of the INS/GNSS-integrated navigation system in the presence of abnormal measurements. On this basis, fully considering the high-order moments of estimation errors, the maximum versoria criterion is introduced as the optimization criterion to construct a novel cost function, further effectively suppressing deviations caused by non-Gaussian disturbances and improving system navigation accuracy. The effectiveness of the proposed method was verified through vehicle navigation experiments. The experimental results demonstrate that the proposed method outperforms traditional approaches, effectively handling measurement anomalies and non-Gaussian measurement noise while maintaining robust navigation performance. Specifically, compared to the EKF, UKF, and MCCUKF, the proposed method reduces the root mean square error of velocity and position by over 60%, 50%, and 30%, respectively, under complex navigation conditions. The algorithm exhibits good accuracy and stability in complex environments, showcasing its practical applicability in real-world navigation systems.