Finite Window Neural Network based Sliding Mode Control for a UAV System

In this paper, we propose a novel finite window neural network-based sliding mode control (FWNN-SMC) for precise and robust control of an unmanned aerial vehicle (UAV) system. The proposed FWNN -SMC defines errors based on the states of the UAV and reference trajectories for each position and attitude and derives control inputs using the sliding mode technique. In the process, to address the nonlinearities of the UAV system and approximate unknown nonlinear functions, we introduce the finite window neural network (FWNN). This guarantees convergence within a finite time and demonstrates robustness to error accumulation and disturbances due to the characteristics of the finite-memory structure.