The essence of INS/GPS navigation is nonlinear. To deal with the accuracy of INS/GPS navigation under
nonlinear, a new nonlinear filtering method, cubature Kalman filter(CKF), is applied to the INS/GPS navigation. Therefor, the
nonlinear state model based on the platform misalignment angle and the observation model described by the velocity error and
position error is established, the CKF filtering principle is analyzed and the INS/GPS integrated filter is designed to simulate
nonlinear model. Simulation results show that CKF reduces the estimation errors of attitude, position and speed compared
with extended Kalman filter(EKF), and CKF prefers to deal with the state estimation problem of integrated navigation.