An Improved In-flight Alignment Method Based on Backtracking Navigation for GPS-aided Low Cost SINS With Short Endurance

Xiaokai Wei,Jie Li,Kaiqiang Feng,Debiao Zhang

DOI: https://doi.org/10.1109/lra.2021.3131703

IF: 5.2

2022-01-01

IEEE Robotics and Automation Letters

Abstract:Aiming at the problem that the low-cost Strapdown Inertial Navigation System (SINS)Global Positioning System (GPS) cannot effectively achieve fast in-flight alignment with short-endurance, an improved alignment method based on backtracking navigation is proposed. Firstly, the existing backtracking alignment scheme is refined by fully considering the low-cost inertial sensor biases and correcting the correlated errors, which can effectively suppress the sensor biases that cannot be ignored while making full use of the limited observation information. Secondly, a cubature Kalman filter (CKF) based on Lie group is designed to correct the sensor biases. The corresponding errors encountered during the construction of the observation vectors for alignment are directly compensated by filtering, thus avoiding the traditional schema of coarse alignment followed by fine alignment, and simplifying the alignment process. The effectiveness of the proposed in-flight alignment method is demonstrated via ground vehicle tests. The test results show that, compared with the existing advanced alignment methods, the proposed method has higher alignment accuracy, faster alignment speed and better stability, and is applicable to low-cost SINS for short-endurance alignment.

robotics