Performance Enhancement of Robust Cubature Kalman Filter for GNSS/INS Based on Gaussian Process Quadrature

Bingbo Cui,Xiyuan Chen,Aichen Wang,Xinhua Wei

doi:10.1109/access.2020.2971003

Bingbo Cui, Xiyuan Chen + Show 2 more

Open Access

https://doi.org/10.1109/access.2020.2971003

Copy DOI

Abstract

The sigma-point Kalman filters are generally considered to outperform extended Kalman filter in the application of GNSS/INS, where cubature Kalman filter (CKF) is widely approved because of its rigorous mathematic derivation. In order to improve the robustness of GNSS/INS under GNSS-challenged environment, a robust CKF (RCKF) is developed based on novel sigma-point update framework (NSUF) in our previous work, whereas the efficiency of NSUF is still plagued by the unknown process model uncertainty. In this paper, an enhanced RCKF is proposed based on Gaussian process quadrature (GPQ), where the uncertainty consisted in sigma points transform is processed by GPQ conditioning on the approximated posterior PDF. Experiment result on loosely coupled GNSS/INS demonstrates the superiority of proposed method, where the heading error and roll error are reduced by 60.5% and 37.5% respectively compared with RCKF, and it achieves better position result than GP-CKF under GNSS outage.

Highlights

Global navigation satellite system (GNSS) has been widely approved as an efficient tool for the navigation of land vehicle because of its superiority in all-weather condition and longterm high accuracy
The aim of this paper is to develop a Gaussian process quadrature (GPQ)-enhanced robust CKF (RCKF) for GNSS/inertial navigation system (INS), which can further improve the attitude of GNSS/INS by considering the uncertainties consisted in process model
In order to improve the prediction stage of RCKF which is very important for attitude estimation of GNSS/INS, a GPQ-based novel sigma-point update framework (NSUF) is proposed in the context of sigma points-based moments matching

Summary

INTRODUCTION

Global navigation satellite system (GNSS) has been widely approved as an efficient tool for the navigation of land vehicle because of its superiority in all-weather condition and longterm high accuracy. In order to process the time-varying measurement noise appeared in NSUF for GNSS/INS, a robust filter named as VB-RCKF is proposed in [22] by VOLUME 8, 2020 employing the VB to update the noise covariance. The novelty of this work is that the uncertainty consisted in moments computation of prior PDF is considered and the efficiency of GP model prediction is improved by applying the posterior PDF approximated by the NSUF-based KF update. B. REVIEW OF NSUF-BASED CKF In order to facilitate the following discussion, the discretetime filter model of GNSS/INS are given by xk = f (xk−1) + wk−1. In the GPQ-based model prediction framework, the effect of these terms can be compensated by using the learned instantiated function values partly, which does not need parameter tune and in turn provides better result compared with normal sigma-point Kalman filter without fixing these terms. In order to enhance the attitude estimation of GNSS/INS without increasing the complexity obviously, only the uncertainty of system function transform is handled by GPQ in our algorithm

FIELD TESTS

Findings

CONCLUSION