Adaptive Iterative Learning Trajectory Tracking Control of 2‐DOF Picking Manipulator Considering Input Constraint

The improvement of picking efficiency and fruit recognition ability in picking tasks has effectively promoted the development process of agricultural intelligence. However, manipulator trajectory‐tracking control, as a core technical link for achieving accurate picking, has not received sufficient attention to its importance. To improve the tracking control performance of the manipulator when performing continuous picking operations in an unstructured environment, this article proposes an improved adaptive iterative learning control (IAILC) strategy that combines an adaptive iterative learning updating law and an input constraint auxiliary variable design system, and the dynamic and static performances of the iterative process are gradually enhanced by learning historical picking information. The adaptive iterative learning updating law estimates the total time‐varying disturbance to meet the accuracy requirements of the picking process, and the introduction of the time‐varying learning gain improves the dynamic performance of the iterative process. The risk of adverse effects on the system's control performance from excessive drive torque is reduced by incorporating input‐constraint auxiliary variables. Both simulation and experimental results have confirmed the trajectory‐tracking performance of the manipulator under external disturbances and model uncertainties using the designed IAILC strategy.