The paper considers a robust innovation-based adaptation method for the extended Kalman filter (EKF) applied to an autonomous strapdown inertial navigation system (SINS) using MEMS-IMU measurements under non-Gaussian outliers. A MATLAB/Simulink simulation framework is used to generate single outlier events, burst outlier sequences, and changes in IMU noise spectral characteristics. Baseline, adaptive, and robust-adaptive EKF configurations with robust innovation weighting (Huber, Tukey) are compared. The evaluation relies on navigation accuracy metrics, NIS consistency, filter divergence probability, and recovery time after outliers in a Monte Carlo experiment series. In addition, the results are interpreted from an operational-economic viewpoint: reducing divergence rate and shortening recovery time are treated as factors that lower mission re-run costs, downtime, and abnormal-mode losses for unmanned and robotic platforms. The results are presented as plots and tables and are used to derive practical recommendations for selecting a robust scheme and tuning parameters for typical interference scenarios.
strapdown inertial navigation system, SINS, EKF, extended Kalman filter, innovation-based adaptation, robust filtering, Huber, Tukey, MEMS IMU, non-Gaussian outliers, NIS, filter divergence, MATLAB/Simulink, simulation, Monte Carlo, operational reliability, economic losses, operational costs, UAS efficiency
1. Meleshko V. V., Nesterenko O. I. Besplatformennye inercial'nye navigacionnye sistemy : ucheb. posobie. — Kirovograd : POLIMED-Servis, 2011. — 164 s. — ISBN 978-966-7813-75-8. — Tekst : elektronnyy. — URL: https://cions.kpi.ua/Arhiv/Meleshko/bins1.pdf (data obrascheniya: 26.02.2026).
2. Voskoboynikov Yu. E. Rekurrentnoe ocenivanie vektora sostoyaniya dinamicheskih sistem : ucheb. posobie. — Novosibirsk : Izd-vo NGTU, 2014. — 136 s. — ISBN 978-5-7782-2486-5. — Tekst : elektronnyy. — URL: https://znanium.ru/catalog/document?id=143456 (data obrascheniya: 26.02.2026
3. Semushin I. V., Cyganova Yu. V., Zaharov K. V. Ustoychivye algoritmy fil'tracii — obzor i novye rezul'taty dlya sistem sudovozhdeniya // Informacionnye tehnologii i vychislitel'nye sistemy. — 2013. — № 4. — S. 90–112. — Tekst : elektronnyy. — URL: https://pzs.dstu.dp.ua/DataMining/kalman/bibl/90_112.pdf (data obrascheniya: 26.02.2026).
4. Stepanov O. A. Fil'tr Kalmana. Istoriya i sovremennost' (k 80-letiyu Rudol'fa Emilya Kalmana) : elektronnyy resurs. — Tekst : elektronnyy. — URL: https://is.ifmo.ru/doctors/kalman-history.pdf (data obrascheniya: 26.02.2026).
5. Sokolov O. A., Pentela O. A., Klevanovich A. S. Adaptivnyy algoritm kompensacii dreyfa besplatformennoy inercial'noy navigacionnoy sistemy v usloviyah avtonomnoy navigacii: povyshenie tochnosti i ekonomicheskaya effektivnost' primeneniya nizkobyudzhetnyh MEMS-datchikov // Zhurnal prikladnyh issledovaniy. — 2025. — № 11. — S. 193–206. — DOI:https://doi.org/10.26118/6644.2025.10.84.027. — Tekst : elektronnyy. — URL: https://jomeam.ru/ru/nauka/article/110366/view (data obrascheniya: 26.02.2026).
6. Materialy MAI po kompleksirovaniyu BINS/SNS i primeneniyu fil'tra Kalmana : elektronnyy resurs (PDF). — Tekst : elektronnyy. — URL: https://mai.ru/upload/iblock/35f/35fe336f42e485af52c8333aa1d1b4b6.pdf (data obrascheniya: 26.02.2026).
7. Kalman R. E. A New Approach to Linear Filtering and Prediction Problems // Transactions of the ASME — Journal of Basic Engineering. — 1960. — Vol. 82, no. 1. — P. 35–45. — DOI:https://doi.org/10.1115/1.3662552. — Tekst : elektronnyy. — URL: https://www.unitedthc.com/DSP/Kalman1960.pdf (data obrascheniya: 26.02.2026).
8. Jazwinski A. H. Adaptive filtering to prevent divergence observed in application of Kalman filter to orbit determination : elektronnyy resurs // NASA Technical Reports Server (NTRS). — 1969. — Document ID: 19690064457. — Tekst : elektronnyy. — URL: https://ntrs.nasa.gov/citations/19690064457 (data obrascheniya: 26.02.2026).
