Research On The Models Of GPS/INS Initial Alignment

Inertial navigation system is a completely autonomous navigation and positioning system, but before starting the navigation and positioning to complete the initial alignment is required to gain a carrier’s priori information. With the development of optical fiber gyro and MEMS inertial components technology, in greatly expand INS applications in the field of civil also brought new challenges to the initial alignment. Paper researches on the model and method for INS initial alignment, including a quaternion based INS navigation recursion arithmetic, static alignment method improvement based on the inertial condensation coordinate system and IGG- Ⅲ equivalent weight function, an INS random misalignment Angle nonlinear error model based on euler misalignment Angle, nonlinear filtering algorithm for nonlinear INS error model and based on GPS location aided INS dynamic alignment. The main research results are as follows:(1)An INS navigation recursion arithmetic based on quaternion was deduced, Attitude to update the navigation precision is crucial, and corresponding to the euler angle and the direction cosine matrix, the quaternion has the advantages of smooth interpolation between bearing, angular displacement sequence and a single angular displacement conversion, and the use of quaternions conjugate can effectively calculate the versa angular displacement, and quaternions only four parameters, low computational burden.(2) Based on the inertial condensation coordinate system inertia condensation coarse alignment method was proposed, the traditional analytic coarse alignment is sensitive to shake, inertial condensation coarse alignment can track angle motion and line movement of vehicle to isolate the effect of shaking; a robust static fine alignment algorithm based on the IGG-Ⅲ equivalent weight function was deduced, the IGG-Ⅲ equivalent weight function can effectively weaken the disturbance of big error of the filtering system.(3) The INS nonlinear error model under random misalignment angle has been set up, its simplified form under the small misalignment Angle and the simulation experiment proves the validity of the model; against the performance of 3th-CKF is not high for INS initial alignment with large misalignment angle, a 5th-CKF was deduced, the contrast test results under different misalignment angle and different nonlinear filtering algorithm show that the 5 CKF has better performance.(4)A GPS aided INS dynamic alignment model based on GPS’s location or speed was deduced, and analyzed the different carrier motions influence on attitude estimation, the simulation experiment results show that the vehicle turning after a certain linear velocity can obtain a better alignment result. The GPS aided INS dynamic alignment method is verified by a set of measured data collected by the SPAN-CPT.