Adaptive Predefined-Time Disturbance Observer-Based Fast Nonsingular Sliding Mode Control Strategy for Consumer Quadrotor UAVs: Theory and Experiments

The assurance of a predefined-time convergence, strong robustness and global adaptability is of great significance for enhancing the control performance and operational reliability of consumer quadrotor unmanned aerial vehicles (UAVs). These properties are particularly valuable for trajectory tracking control systems of UAVs, enabling precise and reliable flight in the presence of unknown disturbances. To address these challenges, this study proposes an adaptive predefined-time disturbance observer (APTDO)-based fast nonsingular sliding mode control strategy. Compared to existing predefined-time stability criterion, a modified predefined-time stability criterion is first developed by incorporating an exponential term and adjustable parameters, which accelerates the convergence speed and allows the actual convergence time to be flexibly adjusted. Building on this criterion, an APTDO is designed to compensate for unknown disturbances. The proposed observer does not require the strict assumption that disturbances are continuously differentiable or negligibly varying, achieves chattering-free estimation, and updates observer gain automatically, thereby enhancing estimation quality and adaptability across diverse missions. Furthermore, the control law integrates a non-monotonic adaptive mechanism and a continuous piecewise function to estimate the upper bound of the disturbance estimation error instead of the disturbance itself, which helps to reduce energy loss and avoid singularity issues. From the Lyapunov analysis, it infers that the control error converges to a sufficiently small region within a predefined time. Ultimately, extensive comparative experiments validate the superiority of the proposed method, highlighting its value and potential for real-world deployment in consumer UAV applications where accuracy, robustness, and reliability are essential for improving system robustness and operational efficiency.