Nonlinear Signal-Correction Observer and Application to UAV Navigation

Abstract

In this paper, a nonlinear signal-correction observer (NSCO) is presented for signals correction and estimation, which not only can reject the position measurement error, but also the unknown velocity can be estimated, in spite of the existence of large position measurement error and intense stochastic non-Gaussian noise. For this method, the position signal is not required to be bounded. The NSCO is developed for position/acceleration integration, and it is applied to an unmanned aerial vehicle (UAV) navigation: Based on the NSCO, the position and flying velocity of quadrotor UAV are estimated. An experiment is conducted to demonstrate the effectiveness of the proposed method.