Research On Attitude Errors Evaluation And Compensation Methods Of SINS Under Large Angular Rate Maneuver Environments

In researches of attitude updating algorithms of strapdown inertial navigation system, Musoff used Jacobian ellipse function to test the performance of the optimal coning algorithms. He verified that when the platform is under coning motion which can be depicted by Jacobian ellipse function, the optimal coning algorithm of Miller is not optimal any more. Even though Musoff claimed that the algorithm of Miller is not optimal under the generalized defined coning motion, he did not give how to obtain the optimal coefficients under this severe coning motion. Based on the coning motion and rotation coexisting environment, performances of traditional coning algorithms are analyzed and evaluated, and higher accuracy attitude updating algorithm is proposed.The contents and researches of the dissertation are mainly about:(a)Normalized quaternions are used for theoretical analysis on the precision of two coning algorithms, frequency-series coning algorithm and explicit frequency shaping coning algorithm, for coning and large angular rate rotation coexisting environments. In this dissertation, second-order approximation is presented and the theoretical precision values improve substantially. The simulations indicate that the coning algorithm based on explicit frequency shaping is more accurate than the one based on frequency series for high-frequency condition under coning and large angular rate rotation coexisting environments.(b) Under coning and angular rate rotation coexisting environments, after theoretical analysis on deducing the value of noncommutativity high order term ignored in traditional coning algorithms, it proved that the value cannot be ignored. High precision algorithm of equivalent rotation vector is designed. Coefficient equations under usual conditions are built and resolved with three-sampling equivalent rotating vector method. Ultimately, based on normalized quaternion and Savage's large angular rate rotation environments, high-order algorithm is designed and simulated. The results of the simulations indicate that the new high-order algorithm can improve the value of attitude update precision.(c) Based on the characteristics of gyroscopes sensors, like scale factor,non-orthogonal error, gyro constant drift etc., noncommutativity errors are fully analyzed. The error equations of ignoring and not ignoring the item are deduced.Analysis show that common quantification method of impulse data can effect the inherent characteristics of signals in usual strapdown inertial navigation system,and FIR filter is designed to decrease the quantification error.