An Adaptive Fault-Tolerant Federated Kalman Filter for a Multi-Sensor Integrated Navigation System.

Area of Science:

• Aerospace Engineering
• Navigation Systems
• Control Theory

Background:

• Autonomous aerospace navigation requires robust systems for all-weather conditions.
• Integrated navigation systems (INS/SRNS/CNS) are susceptible to measurement outliers.
• Existing methods struggle with both slow and abrupt outliers.

Purpose of the Study:

• To develop an adaptive fault-tolerant federated Kalman filter (AFTFKF).
• To enhance the robustness of INS/SRNS/CNS integrated navigation systems against measurement outliers.
• To ensure reliable all-weather cross-domain aerospace navigation.

Main Methods:

• Introduced a noise estimator using maximum likelihood estimation (MLE) and sequential probability ratio test (SPRT) for slow outliers.
• Designed a double residual-based Chi-square test (DCST) information factor for abrupt outliers.
• Integrated SPRT-MLE and DCST into a federated Kalman filter framework to create AFTFKF.

Main Results:

• The proposed AFTFKF demonstrated superior accuracy in simulations.
• The AFTFKF showed strong stability in the presence of measurement outliers.
• Effective mitigation of both slow-growing and abrupt outliers was achieved.

Conclusions:

• The AFTFKF provides a robust solution for integrated aerospace navigation.
• The method enhances system reliability under adverse measurement conditions.
• Achieved autonomous and reliable all-weather cross-domain navigation capabilities.