Investigation into the nonlinear Kalman filter to correct the INS/GNSS integrated navigation system

The integrated navigation system is the inertial navigation system (INS), corrected by global navigation satellite system (GNSS) data. The correction could be done algorithmically by utilizing nonlinear Kalman filtering (NKF). In practice, the NKF uses an INS error model as an a priori model that is not always adequate to handle the dynamics of the true and unknown INS error model. To eliminate such modeling errors, we propose a new INS/GPS correction approach with modified adaptive NKF. In the proposed NKF, instead of the a priori model, the model constructed during the pre-flight test for a particular INS is used. To realize this, the full algorithm includes an INS error model construction algorithm, a way of reduced measurement generation, and criteria for divergence detection. INS error model construction both during pre-flight test and during flight is done by the group method of data handling (GMDH). Flight experiments were performed for an empirical study of the INS error model and its effect on the total accuracy of computed navigational data. The navigational equipment was installed on the balloon—an airborne radio-transparent object. The results of the experiments validate the effectiveness and accuracy of the proposed INS/GPS correction approach.