Observer-Based Adaptive Neural Control of Quadrotor Unmanned Aerial Vehicles Subject to Model Uncertainties and External Disturbances

Quadrotor unmanned aerial vehicles (QUAVs) are widely recognized for their versatility and advantages across diverse applications. However, their inherent instability and underactuated dynamics pose significant challenges, particularly under external disturbances and parametric model uncertainties. This paper presents an advanced observer-based control framework to address these challenges, introducing a high-gain disturbance observer (HGDO) integrated with a neural-network-based adaptive fractional sliding mode control (NN-AFSMC) scheme. The proposed HGDO-NN-AFSMC ensures robust position and attitude tracking by effectively compensating for external disturbances and model uncertainties. A direct control approach is employed, significantly reducing computational complexity by minimizing the need for frequent online parameter updates while maintaining high tracking precision and robustness. The stability of the control system is rigorously analyzed using Lyapunov theory, and comprehensive simulation studies validate the proposed scheme’s superior performance compared to other advanced control approaches, particularly in dynamic and uncertain operational environments. The proposed HGDO-NN-AFSMC achieves a position tracking error of less than 0.03 m and an attitude tracking error below 0.02 radians, even under external disturbances and parametric uncertainties of 20%. Compared to conventional robust feedback linearization (RFBL) and nonsingular fast terminal sliding mode control (NFTSMC), the proposed method improves position tracking accuracy by 25% and reduces settling time by approximately 18%.