Nonlinear Adaptive Control and Reinforcement Learning-Based Anti-Disturbance Trajectory Planning for Variable-Length Cable Suspension Systems

This paper proposes a nonlinear control strategy integrated with variable cable-length adjustment to achieve simultaneous unmanned Aerial Vehicle (UAV) positioning and load swing suppression in aerial transportation systems. For the underactuated system, a dynamic model incorporating timevarying cable length effects and swing angle constraints is established. A Lyapunov-based controller is designed with online load mass estimation through coupled cable-length and altitude tracking errors. Energy-based stability analysis demonstrates the system’s convergence. Reinforcement learning is further integrated to generate anti-swing trajectories, enhancing the system’s adaptability to external disturbances. Simulation experiments verify that the proposed method enables precise UAV positioning, effectively suppresses load swing, and demonstrates robust disturbance rejection capabilities.