Nonlinear Sliding Mode State Estimation Techniques for UAV Application

Abstract

Unmanned air vehicles (UAVs) have received tremendous importance because of their civil, military, industrial, agricultural and security etc. applications. As compared to terrestrial robots, UAVs have to face stronger aerodynamic as well as gravitational forces which effect the dynamics of these flying structures. The control and estimation of unavailable plant states in UAVs is a challenging task due to their nonlinear dynamics in the presence of external uncertainties like air load and noise effects. Practically, external aerodynamic disturbances and sensor noise can alter the desired performance of the aerial robots leading to instability and loss of control of UAVs. Information of all plant states is important for better control and navigation of UAVs. In this paper it is assumed that all plant states are not available, therefore, nonlinear observer design techniques are used to estimate unavailable states of quadrotor plant with least error. The designed nonlinear observer’s results are compared in simulations in the presence of external disturbance and sensor noise. Sliding mode observers implemented in this paper exhibit the property of finite time convergence of error dynamics both with and without disturbances and noise.