A novel hybrid observation prediction methodology for bridging GNSS outages in INS/GNSS systems. (23rd September 2022)

• Record Type: Journal Article
• Authors: Chen, Linzhouting; Liu, Zhanchao; Fang, Jiancheng
• Is Part Of: Journal of navigation. Volume 75:Number 5(2022)
• Page Start: 1206
• Page End: 1225
• Publication Date: 2022-09-23
• Subjects: INS/GNSS -- GNSS outages -- observation prediction -- hybrid PLSR/GPR model -- HP filter; Navigation -- Periodicals; 623.8905
• DOI: 10.1017/S037346332200025X
• Languages: English
• ISSNs: 0373-4633

Abstract

Abstract: The integration of the inertial navigation system (INS) and global navigation satellite system (GNSS) is suited for localisation and navigation applications, such as aircrafts, land vehicles and ships. The primary challenge is for navigation system to achieve accurate and reliable navigation solution during GNSS outages. This paper presents an observation prediction methodology for INS/GNSS bridging GNSS outages, which combines partial least squares regression (PLSR) and Gaussian process regression (GPR) to model the INS/GNSS observations and enable a Kalman filter to estimate INS errors. The performance of proposed PLSR/GPR prediction methodology was validated through four GNSS outages taken on flight experiment data, including diverse manoeuvre conditions. The experiment results demonstrate that remarkable performance enhancements are achieved through applying the proposed PLSR/GPR prediction methodology into INS/GNSS integration.