|
|
Robust GNSS/SINS positioning based on the
SE
2(3)-EKF framework
LI Xin, MENG Shuolin, HUANG Guanwen, ZHANG Qin, LI Hanxu
Acta Geodaetica et Cartographica Sinica
2023, 52 (10):
1640-1649.
DOI: 10.11947/j.AGCS.2023.20220526
For GNSS/SINS integrated navigation, the large errors, such as attitude misalignment angle, will cause the inconsistent coordinates definition of state error and large linearization error, thus the performance of traditional filtering and positioning is reduced, especially in the complex GNSS observation environment. In this paper, the attitude, velocity, and position states are reconstructed as a special
SE
2(3) group element, considering the bias of gyro and accelerometer, a group-vector mixed error model is formed, and then one GNSS/SINS robust filtering algorithm (RLIEKF) based on left invariant measurement is studied. The superiority of the proposed method is validated via the vehicle integrated navigation experiment with large misalignment angle error and GNSS outliers in urban environment. The experimental results show that, compared with traditional EKF method, the attitude angle error is considered in time update and GNSS measurement update of the proposed RLIEKF, thus it has a fast convergence speed under different large misalignment angles, without complicated and long-time attitude alignment steps, which can better deal with the problem such the interrupt GNSS signal during a short time. Because the accuracy of innovation is significantly improved, thus it is more robust to complex observation environment, and with a fast computational efficiency, therefore it has excellent engineering practical value.
Reference |
Related Articles |
Metrics
|
|
