The robust residual-based adaptive estimation Kalman filter method for strap-down inertial and geomagnetic tightly...

Area of Science:

• Navigation Systems
• Signal Processing
• Control Theory

Background:

• Strap-down inertial navigation system/geomagnetic navigation system (SINS/GNS) tightly integrated systems face accuracy degradation due to unknown noise characteristics and measurement outliers.
• Severe cases of outlier presence can lead to filter divergence, compromising navigation reliability.

Purpose of the Study:

• To propose a robust residual-based adaptive estimation Kalman filter (RRAEKF) to enhance the accuracy and stability of SINS/GNS tightly integrated navigation systems.
• To address the challenges posed by unknown noise statistics and outliers in measurement information.

Main Methods:

• The RRAEKF employs covariance matching to detect system abnormalities.
• A weighted factor is introduced to identify and mitigate the impact of outliers in measurement data.
• A contraction factor adaptively adjusts the filter's gain matrix for optimal state and covariance estimation.

Main Results:

• The RRAEKF method demonstrated significant improvements in reducing space position errors for SINS/GNS tightly integrated navigation systems.
• Compared to the standard extended Kalman filter, errors were reduced by 63.37%.
• Compared to the residual-based adaptive estimation method, errors were reduced by 56.93% under time-varying noise and outlier conditions.

Conclusions:

• The proposed RRAEKF method offers a robust solution for improving the filtering accuracy of SINS/GNS tightly integrated navigation systems.
• The RRAEKF effectively handles time-varying noise and outliers, preventing filter divergence and enhancing navigation performance.