A Novel Method for AI-Assisted INS/GNSS Navigation System Based on CNN-GRU and CKF during GNSS Outage

Abstract

In the fields of positioning and navigation, the integrated inertial navigation system (INS)/global navigation satellite systems (GNSS) are frequently employed. Currently, high-precision INS typically utilizes fiber optic gyroscopes (FOGs) and quartz flexural accelerometers (QFAs) rather than MEMS sensors. But when GNSS signals are not available, the errors of high-precision INS also disperse rapidly, similar to MEMS-INS when GNSS signals would be unavailable for a long time, leading to a serious degradation of the navigation accuracy. This paper presents a new AI-assisted method for the integrated high-precision INS/GNSS navigation system. The position increments during GNSS outage are predicted by the convolutional neural network-gated recurrent unit (CNN-GRU). In the process, the CNN is utilized to quickly extract the multi-dimensional sequence features, and GRU is used to model the time series. In addition, a new real-time training strategy is proposed for practical application scenarios, where the duration of the GNSS outage time and the motion state information of the vehicle are taken into account in the training strategy. The real road test results verify that the proposed algorithm has the advantages of high prediction accuracy and high training efficiency.