Comparing between the performance of SVSF with EKF and NH∞ for the autonomous airborne navigation problem

The Nonlinear H∞ (NH∞) has been used to estimate airborne position under the uncertainty of parameter and modeling errors. The min-max solution of the Hœ filter can be considered as a compromise between the optimality and robustness when upper bounds are accurately determined, however, this is not possible in reality. Substantial progress is being made in the field of state estimation, where variable structure control theory and system theory have been used to develop the Smooth Variable Structure Filter (SVSF). The SVSF offers the advantage of robustness to bounded uncertainties and optimal operation of the Kalman Alter. In this paper we studied the effectiveness and robustness of nonlinear SVSF compared to the Extended Kalman Filter (EKF) and NHœ in the presence of the unknown disturbance and noises applied for resolving the unmanned aerial vehicle (UAV) localization problems through Strapdown Inertial Navigation System and Global Positioning System (SINS/GPS) sensor fusion. The simulation results of the implemented filters for the localization problem are given by comparing the true estimation error between the three Alters - EKF, NH∞ - and nonlinear SVSF. Better results of robustness and accuracy are obtained with the nonlinear SVSF filter, which doesn't require any model linearization.