Seamless global positioning system/inertial navigation system navigation method based on square-root cubature Kalman...

Area of Science:

• Navigation Systems Engineering
• Signal Processing
• Machine Learning

Background:

• Global Positioning System (GPS) and Inertial Navigation System (INS) integration is crucial for accurate positioning.
• GPS signal outages significantly degrade the performance of integrated navigation systems.
• Existing methods often struggle to maintain accuracy during prolonged GPS signal loss.

Purpose of the Study:

• To develop a seamless navigation dual-model for robust GPS/INS integration.
• To enhance navigation accuracy and reliability during GPS signal outages.
• To outperform traditional single-model approaches in correcting INS errors.

Main Methods:

• A dual-model approach combining Square-Root Cubature Kalman Filter (SRCKF) and Random Forest Regression (RFR).
• Sub-model 1: Relates INS specific force to filter measurements.
• Sub-model 2: Relates SRCKF cubature points and innovation to filter-induced errors.

Main Results:

• The proposed dual-model ensures seamless navigation capabilities even with complete GPS signal loss.
• Experimental results demonstrate significant improvements in navigation accuracy compared to traditional methods.
• The SRCKF and RFR combination shows superior performance in predicting INS errors.

Conclusions:

• The developed dual-model effectively corrects standalone INS during GPS outages, outperforming single models.
• The integration of SRCKF and RFR provides a more accurate prediction of INS errors.
• This seamless navigation method offers a significant advancement for navigation systems operating in challenging environments.