关键词: IMU, 初始对准, 动基座, 扩展卡尔曼滤波, 组合导航

Abstract: Aiming at the problem that the traditional optimal estimation alignment algorithm (OBA) can not be applied to low-cost INS, this paper proposes an improved OBA algorithm combined with extended Kalman filtering, which realizes the estimation of attitude error with the assistance of GNSS and can be applied to low-cost SINS system. Firstly, the reference vector and observation vector are reconstructed, and the accumulation of bias is suppressed by a fixed integration interval. Secondly, by establishing the correlation between the bias of the gyroscope and the attitude error, a new system state equation is constructed, and the measurement equation is constructed based on the position and velocity of GNSS output. At the same time, the bias and attitude error of the gyro are estimated, and the attitude error is fed back into the construction of multiple vectors, and finally are verified in the simulation experiment and the actual sports car experiment. The experimental results show that compared with the traditional OBA algorithm, the algorithm can realize fast alignment of the moving base, estimate the bias of the gyro and compensate the misalignment angle, improve the alignment accuracy when the carrier is in a long-term motion state, and improve the convergence speed without using sliding window integration. Compared with the FIMA algorithm, the accuracy of the three attitude angles is improved by 47%, 47% and 51%, respectively.

Key words: IMU, initial alignment, moving pedestal, extended Kalman filtering, combined navigation