The SINS' SRC-KF Attitude Estimation Modeling Algorithm

We propose the spherical-radial cubature integral Kalman filtering (SRC-KF) algorithm based on the high-precision requirements for oSINS' attitude determination on aircraft according to Bayesian optimal estimation theory. We obtain the 2n spherical-radial cubature sampling points and their designed sampling points' weights to achieve optimal estimation of the system state vector, thier parameters and their variance matrix through the Gauss-Hermite Quadrature numerical approximation method and the state vectors coordinates transformation from Cartesian coordinate system to spherical coordinate frame, the calculation precision of which can be up to the third order. Using the attitude quaternion method we constructed a new SINS attitude determination nonlinear error model, whose system noise vector depends on system state vector. Meanwhile we constructed a measurement equation whose measurement noise vector depends on quaternion measurement vector by the pseudo observation vector method, and calculated the weighted average of estimated quaternion with Lagrangian operator, and carried out the system noise variance calculation with the system noise variance separation algorithm we designed. Finally we conducted the SINS attitude estimation SRC-KF algorithm simulation on the SINS/CCD attitude estimation experiment platform. It can be seen that the SRC-KF algorithm calculation accuracy was higher than others and has better numerical stability through comparison of the CDKF and UKF algorithms; verifying the feasibility and calculation accuracy of SRC-KF algorithm.