Ultra-Local Model-Based Finite-Time Sliding Mode Control Using Neural Network Observer for Quadrotor Position and Attitude

In this paper, an ultra-local model-based finite-time sliding mode control (ULM-FTSMC) method is developed for tracking quadrotor position and attitude in the presence of uncertainties and external disturbances. Based on an ultra-local model technique, the proposed ULM-FTSMC scheme consists of an adaptive neural network observer (ANNO) and a non-singular fast terminal sliding mode controller (NFTSMC). The ultra-local model is employed to approximate complex quadrotor dynamics, thereby reducing the complexity of controller design. The ANNO is designed to estimate the state variables required for subsequent control design and compensate for the lumped disturbances. Furthermore, an improved reaching law incorporating a variable exponent and multiple power terms is developed for the nonsingular fast terminal sliding surface, based on which an NFTSMC is constructed to achieve accurate trajectory tracking within finite time. The stability of the closed-loop system and the finite-time convergence of the tracking errors are rigorously established using Lyapunov theory. Finally, comparative numerical simulations with several existing controllers are conducted to demonstrate the effectiveness and superiority of the proposed method.