An Intelligent Robust Control Strategy for an Uncertain Quadrotor

This paper presents an intelligent robust controller design scheme for a quadrotor where system dynamics of quadrotor are not known and quadrotor actuators may suffer from partial loss of effectiveness (PLOE) along with the actuator saturation. Robust controller design employs an integral fast terminal sliding mode controller (IFTSMC) which provides faster response and singularity free control action along with the finite-time convergence. Actuator saturation is accommodated by employing an auxiliary variable. To cope with the unknown quadrotor dynamics, a radial basis function neural network (RBFN) along with the robust compensator has been introduced to directly approximate control law instead of approximating system dynamics separately. The letter requires more computational effort and that is greatly reduced by the proposed methodology. Weights of RBFN are updated in real-time where update laws of these weights are obtained using the principle of Lyapunov stability and overall stability of the system is guaranteed using the same. To show the effectiveness of the proposed approach, simulation results have been provided for ∞-shaped trajectory tracking.