Litcius/Paper detail

Novel Augmented Quaternion UKF for Enhanced Loosely Coupled GPS/INS Integration

Ahmed M. Elsergany, Mamoun F. Abdel–Hafez, Mohammad A. Jaradat

2024IEEE Transactions on Control Systems Technology17 citationsDOIOpen Access PDF

Abstract

This article presents a novel direct filtering approach for loosely coupled global positioning system (GPS) and inertial navigation system (INS) integration. The proposed model is established based on utilizing the full nonlinear INS state equations in a direct configuration while including vehicle orientation through a unit-quaternion representation. A novel augmented quaternion unscented Kalman filter (AQUKF) is developed and proposed to address the direct nonlinear estimation of vehicle states for outdoor vehicle localization while preserving the non-Euclidean geometry of unit-quaternions. The proposed filter is experimentally validated under full GPS coverage as well as prolonged GPS outages. Results obtained in this article show that the proposed filter outperforms other existing solutions in various experimental testing scenarios.

Topics & Concepts

QuaternionGlobal Positioning SystemGPS/INSInertial navigation systemKalman filterComputer scienceNonlinear systemControl theory (sociology)Inertial measurement unitOrientation (vector space)Filter (signal processing)Extended Kalman filterGPS signalsAssisted GPSAlgorithmArtificial intelligenceComputer visionMathematicsControl (management)TelecommunicationsGeometryPhysicsQuantum mechanicsInertial Sensor and NavigationIndoor and Outdoor Localization TechnologiesGNSS positioning and interference