Research On Initial Alignment Technology Of Strapdown Inertial Navigation

Under the constraints of take-off weight, cost, performance and other factors in Unmanned Aerial Vehicle (UAV), MEMS inertial navigation equipment has been widely applied. And yet rapidly and accurately initial alignment technology in MEMS inertial navigation equipment is the key issues. Therefore the initial alignment of MEMS-INS is researched in this paper, in which improved Kalman filter model for fine alignment under static conditions is designed,and then initial alignment test is carried out. Aiming at transfer alignment problem for carrier-based UAV, a rapid and effective speed plus measuring attitude quaternion matching approach is proposed.In the process of MEMS-INS initial alignment, due to the low accuracy of MEMS gyroscope, auxiliary information is integrated to improve the alignment accuracy. In this paper, using MIMU and MEMS magnetometer information, MEMS-INS error model is built, and thus integrated Kalman filter is designed. Inertial measurement system of AHRS440from Crossbow is applied as the IMU. The EMU data, including acceleration, angular velocity, and attitude data, are received through serial port communication, and then parsed and stored as the original data of the alignmet in accordance with the data protocol of AHRS440. Next, the designed alignment method is implemented to estimate the misalignment angles and sensors’ features. Finally,the results showthat the proposed method is feasible.Also,aiming at UAV transfer alignment problem in mobile base, a rapid speed plus measuring attitude quaternion matching approach is proposed, based on speed matching and speed plus attitude matching methods. On the basis of matching methods, the mechanism and compensation methods of lever-arm effect error and flexibility distortion error are studied respectively. For compensating lever-arm effect error, two methods, digital low-pass filer and direct computing, are adopted. For compensating flexibility distortion, second order Markov process is applied to model flexibility distortion error,and then related state variables to it are induced to system state equations. Finally, using designed transfer alignment simulation system based on Matlab, the proposed transfer alignment approach is simulated. The simulation results show that the proposed speed and measuring attitude quaternion method has high estimation precision and strong resistance to the bad influence of the lever-arm effect error and flexure deformation on the process of transfer alignment.