Antiswing Control of Underactuated Rotary Cranes: An Improved Iterative Learning Scheme

Rotary cranes are critical in industries such as manufacturing, military, and construction, where high safety and control performance standards are required due to complex operating environments. While existing automatic control methods have advanced in tracking and swing suppression, they often overlook valuable insights from historical data and the practical limitations of motor output torque, which complicate implementation and hinder performance. This article proposes an improved iterative learning control scheme that utilizes nonstrict repetition historical data to enhance tracking accuracy and suppress swing, even under input saturation and uncertainties. Specifically, the proposed approach incorporates refined trajectory adjustment and time scale transformation, improving the selfregulation of the control strategy in nonstrict repetition iterations. Furthermore, a smoothing function is introduced to address the nonlinear effects of input saturation, simplifying controller design. The finite-time convergence of the closed-loop system is rigorously proven using a composite energy function. Simulation and experimental results demonstrate that the proposed method effectively reduces payload swings and ensures precise tracking, validating its effectiveness.