An integration algorithm for SINS/GNSS/Airdata navigation system using adaptive super-twisting method + optimized...

Area of Science:

• Navigation Systems Engineering
• Control Systems Theory
• Machine Learning for Sensor Fusion

Background:

• Integrated navigation systems rely heavily on Global Navigation Satellite System (GNSS) data.
• Loss of GNSS signals due to environmental factors or jamming severely degrades navigation accuracy.
• Strap-down Inertial Navigation Systems (SINS) are susceptible to accumulated errors and unmodeled inertial measurement biases.

Purpose of the Study:

• To design and implement a robust integrated navigation system resilient to Global Navigation Satellite System (GNSS) signal outages.
• To enhance navigation accuracy by addressing Strap-down Inertial Navigation System (SINS) perturbations and inertial measurement errors.
• To leverage an optimized hybrid GRU-LSTM sequential model (OHGLSM) within an adaptive super-twisting observer framework.

Main Methods:

• An adaptive super-twisting observer is employed for state and perturbation estimation.
• An optimized hybrid GRU-LSTM sequential model (OHGLSM) is integrated for enhanced position estimation.
• Air-data sensors (pitot tube, barometric pressure sensor) provide velocity and altitude data, utilizing non-holonomic constraints.
• Extra states are incorporated into the nominal model to account for unknown inertial measurement errors.

Main Results:

• The proposed method demonstrates high accuracy and reliability in integrated navigation, even during Global Navigation Satellite System (GNSS) signal blockages.
• Experimental evaluations using real-world vehicle tests validate the system's performance across various scenarios.
• Comparative analyses show superior accuracy compared to existing estimation methods, particularly during GNSS outages.

Conclusions:

• The developed adaptive observer with OHGLSM provides a reliable solution for integrated navigation during Global Navigation Satellite System (GNSS) signal loss.
• The integration of air-data sensors enhances robustness by providing crucial information when GNSS is unavailable.
• The proposed approach significantly improves the precision and reliability of navigation systems in challenging environments.