Research On Strapdown Inertial Initial Alignment Aigorithm Based On Quaternion Optimal Estimation

The initial alignment process is a key technology of strapdown inertial navigation system,its accuracy and computing speed directly determine the performance of strapdown inertial navigation system.Most of the traditional initial alignment methods use a two-step alignment process with coarse alignment and fine alignment,but whether it is coarse or fine alignment,it needs to carry out attitude calculation,because the Euler angle and direction cosine of the two commonly used attitude calculation methods are computationally intensive and inefficient.The quaternion description is widely used in the attitude calculation process of the strapdown inertial navigation system because of its small computation and non-singularity.The initial alignment method of quaternion attitude estimation has become a hotspot.In this paper,based on the existing quaternion optimal estimation alignment method,we in depth analyze the algorithm alignment model and the performance.The initial alignment algorithm based on quaternion optimal estimation is improved and optimized.The research work of this paper is mainly from the following points:Firstly,a quaternion optimal estimation alignment linear model is established by the transformation of the quaternion linear pseudo-metric equation for the quaternion optimal estimation alignment method.Then,the observable performance of the error model is analyzed,and the observability of the alignment model is analyzed qualitatively and quantitatively by using the piecewise linear steady system analysis theory and singular value decomposition theory.It is proved that the alignment model is fully observable.On the basis of this,the quaternion Kalman filter algorithm is used to estimate the initial attitude quaternion in real time,and the accurate initial attitude estimation is obtained.The precise alignment of the one-step estimation is realized.The algorithm is validated by simulation experiments and experimental data.Secondly,the adaptive quaternion Kalman algorithm is proposed to solve the problem of measurement noise caused by gyro drift and accelerometer misalignment error.The adaptive quadratic Kalman filter is proposed to achieve the optimal estimation and alignment.The simulation results and experimental results show that the algorithm can not only improve the estimation of the proposed method.Accuracy,but also improve the system's robustness and real-time.Finally,an optimal estimation alignment algorithm based on matrix Kalman filter is proposed in this paper,which is influenced by gyro drift and zero deviation of accelerometer in the calculation process of K matrix for optimal quaternion estimation alignment.By applying the matrix Kalman filter to the K matrix,the noise effect of the K matrix is considered to some extent,and then the optimal estimation method is used to calculate the more accurate initial quadratic.Simulation and experimental results show that the improved method can effectively improve the initial alignment speed and alignment accuracy...